Tuesday, January 13, 2015

A Brief Overview of Deep Learning

(This is a guest post by Ilya Sutskever on the intuition behind deep learning as well as some very useful practical advice. Many thanks to Ilya for such a heroic effort!)

Deep Learning is really popular these days. Big and small companies are getting into it and making money off it. It’s hot. There is some substance to the hype, too: large deep neural networks achieve the best results on speech recognition, visual object recognition, and several language related tasks, such as machine translation and language modeling.

But why? What’s so special about deep learning? (from now on, we shall use the term Large Deep Neural Networks --- LDNN --- which is what the vaguer term “Deep Learning” mostly refers to). Why does it work now, and how does it differ from neural networks of old? Finally, suppose you want to train an LDNN. Rumor has it that it’s very difficult to do so, that it is “black magic” that requires years of experience. And while it is true that experience helps quite a bit, the amount of “trickery” is surprisingly limited ---- one needs be on the lookout for only a small number well-known pitfalls. Also, there are many open-source implementations of various state-of-the-art neural networks (c.f. Caffe, cuda-covnet, Torch, Theano), which makes it much easier to learn all the details needed to make it work.
Why Does Deep Learning Work?
It is clear that, to solve hard problems, we must use powerful models. This statement is obvious. Indeed, if a model is not powerful, then there is absolutely no chance that it can succeed in solving a hard problem, no matter how good the learning algorithm is.

The other necessary condition for success is that our model is trainable. That too is obvious, for if we cannot train our model, then its power is useless --- it will never amount to anything, and great results will not be achieved. The model will forever remain in a state of unrealized potential.

Fortunately, LDNNs are both trainable and powerful.
Why Are LDNNs Powerful?
When I talk about LDNNs, I’m talking about 10-20 layer neural networks (because this is what can be trained with today’s algorithms). I can provide a few ways of looking at LDNNs that will illuminate the reason they can do as well as they do.

  • Conventional statistical models learn simple patterns or clusters. In contrast, LDNNs learn computation, albeit a massively parallel computation with a modest number of steps. Indeed, this is the key difference between LDNNs and other statistical models.

  • To elaborate further: it is well known that any algorithm can be implemented by an appropriate very deep circuit (with a layer for each timestep of the algorithm’s execution -- one example). What’s more, the deeper the circuit, the more expensive are the algorithms that can be implemented by the circuit (in terms of runtime). And given that neural networks are circuits as well, deeper neural networks can implement algorithms with more steps ---- which is why depth = more power.
    • N.B.: It is easy to see that a single neuron of a neural network can compute the conjunction of its inputs, or the disjunction of its inputs, by simply setting their connections to appropriate values.

  • Surprisingly, neural networks are actually more efficient than boolean circuits. By more efficient, I mean that a fairly shallow DNN can solve problems that require many more layers of boolean circuits. For a specific example, consider the highly surprising fact that a DNN with 2 hidden layer and a modest number of units can sort N N-bit numbers! I found the result shocking when I heard about it, so I implemented a small neural network and trained it to sort 10 6-bit numbers, which was easy to do to my surprise. It is impossible to sort N N-bit numbers with a boolean circuit that has two hidden layers and that are not gigantic.
    • The reason DNNs are more efficient than boolean circuits is because neurons perform a threshold operation, which cannot be done with a tiny boolean circuit.

  • Finally, human neurons are slow yet humans can perform lots of complicated tasks in a fraction of a second. More specifically, it is well-known that a human neuron fires no more than 100 times per second. This means that, if a human can solve a problem in 0.1 seconds, then our neurons have enough time to fire only 10 times --- definitely not much more than that. It therefore follows that a large neural network with 10 layers can do anything a human can in 0.1 seconds.

  • This is not scientific fact since it is conceivable that real neurons are much more powerful than artificial neurons, but real neurons may also turn out to be much less powerful than artificial neurons. In any event, the above is certainly a plausible hypothesis.

  • This is interesting because humans can solve many complicated perception problems in 0.1 seconds --- for example, humans can recognize the identity of an object that’s in front of them, recognize a face, recognize an emotion, and understand speech in a fraction of a second. In fact, if there exists even just one person in the entire world who has achieved an uncanny expertise in performing a highly complex task of some sort in a fraction of a second, then this is highly convincing evidence that a large DNN could solve the same task --- if only its connections are set to the appropriate values.

  • But won’t the neural network need to be huge? Maybe. But we definitely know that it won’t have to be exponentially large ---- simply because the brain isn’t exponentially large! And if human neurons turn out to be noisy (for example), which means that many human neurons are required to implement a single real-valued operation that can be done using just one artificial neuron, then the number of neurons required by our DNNs to match a human after 0.1 seconds is greatly diminished.

These four arguments suggest (strongly, in my opinion), that for a very wide variety of problems, there exists a setting of the connections of a LDNN that basically solves the problem. Crucially, the number of units required to solve these problems is far from exponential --- on the contrary, the number of units required is often so “small” that it is even possible, using current hardware, to train a network that achieves super-high performance on the task of interest. It is this last point which is so important, and requires additional elaboration:

  • We know that most machine learning algorithms are consistent: that is, they will solve the problem given enough data. But consistency generally requires an exponentially large amount of data. For example, the nearest neighbor algorithm can definitely solve any problem by memorizing the correct answer to every conceivable input. The same is true for a support vector machine --- we’d have a support vector for almost every possible training case for very hard problems. The same is also true for a neural network with a single hidden layer: if we have a neuron for every conceivable training case, so that neuron fires for that training case and but not for any other, then we could also learn and represent every conceivable function from inputs to outputs. Everything can be done given exponential resources, but it is never ever going to be relevant in our limited physical universe.

  • And it is in this point that LDNNs differ from previous methods: we can be reasonably certain that a large but not huge LDNN will achieve good results on a surprising variety of problems that we may want to solve. If a problem can be solved by a human in a fraction of a second, then we have a very non-exponential super-pessimistic upper bound on the size of the smallest neural network that can achieve very good performance.

  • But I must admit that it is impossible to predict whether a given problem will be solvable by a deep neural network ahead of time, although it is often possible to tell whenever we know that a similar problem can be solved by an LDNN of a manageable size.

So that’s it, then. Given a problem, such as visual object recognition, all we need is to train a giant convolutional neural network with 50 layers. Clearly a giant convnet with 50 layers can be configured to achieve human-level performance on object recognition --- right? So we simply need to find these weights. Once once we do, the problem is solved.
Learning.
What is learning? Learning is the problem of finding a setting of the neural network’s weights that achieves the best possible results on our training data. In other words, we want to “push” the information from the labelled data into the parameters so that the resulting neural network will solve our problem.

The success of Deep Learning hinges on a very fortunate fact: that well-tuned and carefully-initialized stochastic gradient descent (SGD) can train LDNNs on problems that occur in practice. It is not a trivial fact since the training error of a neural network as a function of its weights is highly non-convex. And when it comes to non-convex optimization, we were taught that all bets are off. Only convex is good, and non-convex is bad. And yet, somehow, SGD seems to be very good at training those large deep neural networks on the tasks that we care about. The problem of training neural networks is NP-hard, and in fact there exists a family of datasets such that the problem of finding the best neural network with three hidden units is NP-hard. And yet, SGD just solves it in practice. This is the main pillar of deep learning.

We can say fairly confidently that successful LDNN training relies on the “easy” correlation in the data, which allows learning to bootstrap itself towards the more “complicated” correlations in the data. I have done an experiment that seems to support this claim: I found that training a neural network to solve the parity problem is hard. I was able to train the network to solve parity for 25 bits, 29 bits, but never for 31 bits (by the way, I am not claiming that learning parity is impossible for over 30 bits --- only that I didn’t succeed in doing so). Now, we know that parity is a highly unstable problem that doesn’t have any linear correlations: every linear function of the inputs is completely uncorrelated with the output, which is a problem for neural networks since they are mostly linear at initialization time (so perhaps I should’ve used larger initial weights? I will discuss the topic of weight initialization later in the text). So my hypothesis (which is shared by many other scientists) is that neural networks start their learning process by noticing the most “blatant” correlations between the input and the output, and once they notice them they introduce several hidden units to detect them, which enables the neural network to see more complicated correlations. Etc. The process goes on. I imagine some sort of a “spectrum” of correlations --- both easy and hard, and the network jumps from a correlation to a more complicated correlation, much like an opportunistic mountain climber.
Generalization.
While it is very difficult to say anything specific about the precise nature of the optimization of neural networks (except near a local minimum where everything becomes convex and uninteresting), we can say something nontrivial and specific about generalization.

And the thing we can say is the following: in his famous 1984 paper called "A Theory of the Learnable", Valiant proved, roughly speaking, that if you have a finite number of functions, say N, then every training error will be close to every test error once you have more than log N training cases by a small constant factor. Clearly, if every training error is close to its test error, then overfitting is basically impossible (overfitting occurs when the gap between the training and the test error is large). (I am also told that this result was given in Vapnik’s book as small exercise). This theorem is easy to prove but I won’t do it here.

But this very simple result has a genuine implication to any implementation of neural networks. Suppose I have a neural network with N parameters. Each parameter will be a float32. So a neural network is specified with 32N bits, which means that we have no more than 232N distinct neural networks, and probably much less. This means that we won’t overfit much once we have more than 32N training cases. Which is nice. It means that it’s theoretically OK to count parameters. What’s more, if we are quite confident that each weight only requires 4 bits (say), and that everything else is just noise, then we can be fairly confident that the number of training cases will be a small constant factor of 4N rather than 32N.
The Conclusion:
If we want to solve a hard problem we probably need a LDNN, which has many parameters. So we need a large high-quality labelled training set to make sure that it has enough information to specify all the network’s connections. And once we get that training set, we should run SGD on it until the network solves the problem. And it probably will, if our neural network is large and deep.
What Changed Since the 80s?
In the old days, people believed that neural networks could “solve everything”. Why couldn’t they do it in the past? There are several reasons.

  • Computers were slow. So the neural networks of past were tiny. And tiny neural networks cannot achieve very high performance on anything. In other words, small neural networks are not powerful.

  • Datasets were small. So even if it was somehow magically possible to train LDNNs, there were no large datasets that had enough information to constrain their numerous parameters. So failure was inevitable.

  • Nobody knew how to train deep nets. Deep networks are important. The current best object recognition networks have between 20 and 25 successive layers of convolutions. A 2 layer neural network cannot do anything good on object recognition. Yet back in the day everyone was very sure that deep nets cannot be trained with SGD, since that would’ve been too good to be true!

It’s funny how science progresses, and how easy it is to train deep neural networks, especially in retrospect.
Practical Advice.
Ok. So you’re sold. You’re convinced that LDNNs are the present and the future and you want to train it. But rumor has it that it’s so hard, so difficult… or is it? The reality is that it used to be hard, but now the community has consolidated its knowledge and realized that training neural networks is easy as long as you keep the following in mind.

Here is a summary of the community’s knowledge of what’s important and what to look after:
  • Get the data: Make sure that you have a high-quality dataset of input-output examples that is large, representative, and has relatively clean labels. Learning is completely impossible without such a dataset.

  • Preprocessing: it is essential to center the data so that its mean is zero and so that the variance of each of its dimensions is one. Sometimes, when the input dimension varies by orders of magnitude, it is better to take the log(1 + x) of that dimension. Basically, it’s important to find a faithful encoding of the input with zero mean and sensibly bounded dimensions. Doing so makes learning work much better. This is the case because the weights are updated by the formula: change in wij \propto xidL/dyj (w denotes the weights from layer x to layer y, and L is the loss function). If the average value of the x’s is large (say, 100), then the weight updates will be very large and correlated, which makes learning bad and slow. Keeping things zero-mean and with small variance simply makes everything work much better.

  • Minibatches: Use minibatches. Modern computers cannot be efficient if you process one training case at a time. It is vastly more efficient to train the network on minibatches of 128 examples, because doing so will result in massively greater throughput. It would actually be nice to use minibatches of size 1, and they would probably result in improved performance and lower overfitting; but the benefit of doing so is outweighed the massive computational gains provided by minibatches. But don’t use very large minibatches because they tend to work less well and overfit more. So the practical recommendation is: use the smaller minibatch that runs efficiently on your machine.

  • Gradient normalization: Divide the gradient by minibatch size. This is a good idea because of the following pleasant property: you won’t need to change the learning rate (not too much, anyway), if you double the minibatch size (or halve it).

  • Learning rate schedule: Start with a normal-sized learning rate (LR) and reduce it towards the end.
    • A typical value of the LR is 0.1. Amazingly, 0.1 is a good value of the learning rate for a large number of neural networks problems. Learning rates frequently tend to be smaller but rarely much larger.
    • Use a validation set ---- a subset of the training set on which we don’t train --- to decide when to lower the learning rate and when to stop training (e.g., when error on the validation set starts to increase).
    • A practical suggestion for a learning rate schedule: if you see that you stopped making progress on the validation set, divide the LR by 2 (or by 5), and keep going. Eventually, the LR will become very small, at which point you will stop your training. Doing so helps ensure that you won’t be (over-)fitting the training data at the detriment of validation performance, which happens easily and often. Also, lowering the LR is important, and the above recipe provides a useful approach to controlling via the validation set.

  • But most importantly, worry about the Learning Rate. One useful idea used by some researchers (e.g., Alex Krizhevsky) is to monitor the ratio between the update norm and the weight norm. This ratio should be at around 10-3. If it is much smaller then learning will probably be too slow, and if it is much larger then learning will be unstable and will probably fail.

  • Weight initialization. Worry about the random initialization of the weights at the start of learning.
    • If you are lazy, it is usually enough to do something like 0.02 * randn(num_params). A value at this scale tends to work surprisingly well over many different problems. Of course, smaller (or larger) values are also worth trying.
    • If it doesn’t work well (say your neural network architecture is unusual and/or very deep), then you should initialize each weight matrix with the init_scale / sqrt(layer_width) * randn. In this case init_scale should be set to 0.1 or 1, or something like that.
    • Random initialization is super important for deep and recurrent nets. If you don’t get it right, then it’ll look like the network doesn’t learn anything at all. But we know that neural networks learn once the conditions are set.
    • Fun story: researchers believed, for many years, that SGD cannot train deep neural networks from random initializations. Every time they would try it, it wouldn’t work. Embarrassingly, they did not succeed because they used the “small random weights” for the initialization, which works great for shallow nets but simply doesn’t work for deep nets at all. When the nets are deep, the many weight matrices all multiply each other, so the effect of a suboptimal scale is amplified.
    • But if your net is shallow, you can afford to be less careful with the random initialization, since SGD will just find a way to fix it.
    You’re now informed. Worry and care about your initialization. Try many different kinds of initialization. This effort will pay off. If the net doesn’t work at all (i.e., never “gets off the ground”), keep applying pressure to the random initialization. It’s the right thing to do.

  • If you are training RNNs or LSTMs, use a hard constraint over the norm of the gradient (remember that the gradient has been divided by batch size). Something like 15 or 5 works well in practice in my own experiments. Take your gradient, divide it by the size of the minibatch, and check if its norm exceeds 15 (or 5). If it does, then shrink it until it is 15 (or 5). This one little trick plays a huge difference in the training of RNNs and LSTMs, where otherwise the exploding gradient can cause learning to fail and force you to use a puny learning rate like 1e-6 which is too small to be useful.

  • Numerical gradient checking: If you are not using Theano or Torch, you’ll be probably implementing your own gradients. It is easy to make a mistake when we implement a gradient, so it is absolutely critical to use numerical gradient checking. Doing so will give you a complete peace of mind and confidence in your code. You will know that you can invest effort in tuning the hyperparameters (such as the learning rate and the initialization) and be sure that your efforts are channeled in the right direction.

  • If you are using LSTMs and you want to train them on problems with very long range dependencies, you should initialize the biases of the forget gates of the LSTMs to large values. By default, the forget gates are the sigmoids of their total input, and when the weights are small, the forget gate is set to 0.5, which is adequate for some but not all problems. This is the one non-obvious caveat about the initialization of the LSTM.

  • Data augmentation: be creative, and find ways to algorithmically increase the number of training cases that are in your disposal. If you have images, then you should translate and rotate them; if you have speech, you should combine clean speech with all types of random noise; etc. Data augmentation is an art (unless you’re dealing with images). Use common sense.

  • Dropout. Dropout provides an easy way to improve performance. It’s trivial to implement and there’s little reason to not do it. Remember to tune the dropout probability, and to not forget to turn off Dropout and to multiply the weights by (namely by 1-dropout probability) at test time. Also, be sure to train the network for longer. Unlike normal training, where the validation error often starts increasing after prolonged training, dropout nets keep getting better and better the longer you train them. So be patient.

  • Ensembling. Train 10 neural networks and average their predictions. It’s a fairly trivial technique that results in easy, sizeable performance improvements. One may be mystified as to why averaging helps so much, but there is a simple reason for the effectiveness of averaging. Suppose that two classifiers have an error rate of 70%. Then, when they agree they are right. But when they disagree, one of them is often right, so now the average prediction will place much more weight on the correct answer. The effect will be especially strong whenever the network is confident when it’s right and unconfident when it’s wrong.

I am pretty sure that I haven’t forgotten anything. The above 13 points cover literally everything that’s needed in order to train LDNNs successfully.
So, to Summarize:
  • LDNNs are powerful.
  • LDNNs are trainable if we have a very fast computer.
  • So if we have a very large high-quality dataset, we can find the best LDNN for the task.
  • Which will solve the problem, or at least come close to solving it.
The End.
But what does the future hold? Predicting the future is obviously hard, but in general, models that do even more computation will probably be very good. The Neural Turing Machine is a very important step in this direction. Other problems include unsupervised learning, which is completely mysterious and incomprehensible in my opinion as of 8 Jan 2015. Learning very complicated “things” from data without supervision would be nice. All these problems require extensive research.

832 comments:

«Oldest   ‹Older   801 – 832 of 832
INVEST WITH $200 AND GET A RETURNS OF $2,000 said...


Invest with 200$ and get a returns of 5,000$ within seven business working days.
Why wasting your precious time online looking for a loan? When there is an opportunity for you to invest with 200$ and get a returns of 5,000$ within seven business working days. Contact us now for more information if interested on how you can earn big with just little amount. This is all about investing into Crude Oil and Gas Business.

Email: HappyInvestment-world_inc@protmail.com

nisha raj said...

Your post is just outstanding! thanx for such a post,its really going great and great work.
python training in kalyan nagar|python training in marathahalli
selenium training in marathahalli|selenium training in bangalore
devops training in kalyan nagar|devops training in bellandur
phthon training in bangalore

robin said...

Good post!Thank you so much for sharing this lovely article.It was so good to read and useful to upgrade my understanding...
Hadoop training in Bangalore|
Big Data Training in Bangalore|
Big Data Analytics Training in Bangalore|
Hadoop Training in Bellandur|
Hadoop Training in Marathahalli|
Angular JS Training in Bangalore

vanitha said...

Thank you for a very good article. You've produced an interesting article.
Lyceum Northwestern University , situated in Dagupan, Philippines, is a world top university. Lyceum Northwestern University-Choose the curriculum, compare tuition fees and apply online.
Good work! Congratulations on the blog!!

Cognex Technology said...

Hey, thanks for sharing these amazing details. keep blogging. For more updates Visit Cognex Technology

jameswill11 said...

Any newbie to QuickBooks will find it difficult in terms of installing the software to their personal computer or laptop, create bills, enter down bank information or update the program to the recently released version. All those difficulties may be solved instantly by reaching off to the executives . QuickBooks Support Phone Number is available 24/7 to give much-needed integration related support.

buy dissertation online said...

Thanks for sharing this post. I liked it because it is so authentic and useful. Your blog is so beautiful. I appreciate you for your hard work.

Quickbooks Enhanced Payroll Customer Support said...

To know more about our services, check out the list of the issues we solve. If your company also uses QuickBooks POS software to manage your sales, inventory and customers’ information, we have experts who can help you maintain this system too. Apart from troubleshooting, we can help you set up custom reporting, assist in migrating your data from old accounting system to QB, provide regular file cleanup and other enhancement services. We offer several support packages: basic, premium, and pro. Your choice of a package will depend on the number of software users, and a time period you want to receive our services. The packages include unlimited QuickBooks Helpline Number and 24/7 technical support. With our assistance, your business operations will never be interrupted due to the software failure.

steffan said...

The QuickBooks Tech Support Phone Number stays top-notch because of the fact technicians that will be planning to deal with your queries are trained to be practical and versatile when it comes to taking quick decisions which will instantly solve the reported error.

unknow said...

zI really enjoyed your blog Thanks for sharing such an informative post.
https://myseokhazana.com/
https://seosagar.in/
Indian Bookmarking list
Indian Bookmarking list
India Classified Submission List
Indian Classified List
Indian Bookmarking list
Indian Bookmarking list
India Classified Submission List
Indian Classified List
Friendship Club In India
Friendship Club In Mumbai
Friendship Club In Delhi
Friendship Club
Friendship Chandigarh

Unknown said...

Extraordinary Article! I truly acknowledge this.You are so wonderful! This issue has and still is so significant and you have tended to it so Informative.
Visit for best website design and seo services at-webdots.in
best web designing company in hubli
best digital marketing companhy in hubli
best web designing company in dharwad
best digital marketing company in dahrwad
best web designing company in shimoga
best digital marketing company in shimoga
best web designing company in belgaum
best digital marketing company in belgaum
best web designing company in davangere
best digital marketing company in davangere

HP Printer Expert said...

Thank you for sharing your blog. To find out more, visit our website.
123.hp.com || 123.hp.com/setup || 123HP Setup || hp.com/setup || hp.com/123 || 123.hp.com setup || 123 HP Printer Setup || 123 HP Printer Support || 123 HP Setup and Install || 123hpcom || 123 HP Printer Install || 123hpcomsetup || 123 HP Wireless Setup || 123 HP Install || hpcom/123 || 123hpcominstall || 123HP Setup || 123 HP Smart App || Install 123 HP Printer || HP 123 Setup Scanner

Anonymous said...

Hi, Your Blog is Amazing. Thank you for sharing.

Visit us for Antivirus Related Queries,

mcafee antivirus scan not working


windows defender


common antivirus error


create bitdefender rescue cd


rescue mode in bitdefender


how to use bitdefender wallet


avg antivirus error


access norton backup files


eset error codes


bitdefender mobile security

Surbhi Singh said...

Very nice and informative blog. I like the information shared by you very much. Similar services are provided by TheTuitionTeacher in Lucknow.
Tuition Service Lucknow | Home Tuition Service

Unknown said...


Many of vape customers delight in Vaping without Nicotine since the tastes that are present in the vape are of excellent quality. They consist of sweet food cravings like cinnamons and also cheesecake. Customers take many calories which include the flavors. This is much more helpful to those that smoke without nicotine. Because of this, lots of vape consumers are fleing from the vape which contains extra pure nicotine to ones which have no Nicotine. They have observed that taking more Nicotine has an addiction that might make it hard for you to manage how much you vape and create prospective wellness risks in the future including cancer cells weed vape pen.
It is one means to rather a Pure nicotine addiction dab pen

Lots of people smoke to satisfy themselves and even to please their social peers as well as close friends and also not really feel neglected. They may try to try to find much healthier cigarettes if they can not stay clear of cigarette smoking. This is because a lot of cigarettes have way too much percentage of Pure nicotine that becomes so addicting making it difficult to give up cigarette smoking and even control the price at which you smoke. To avoid this, you need to get used to vape tastes which include no nicotine. They are not addicting and also contain no prospective health dangers. This will not also impact your need for smoking since you will certainly be currently smoking a healthier vape as well as you will certainly find it very easy to give up when you seem like.phix
Method Vaping

Given that many individuals have actually been addicted to the Pure nicotine vapes, it provides a hard time to ignore what they made use of to enter into vapes which consist of Pure nicotine. They have actually currently progressed in that they are currently utilizing their MODs methods to make their smoke strike a lot easier. You require to obtain much into these methods which will aid you to consume vapes which have less pure nicotine. You can do this by obtaining the no-nicotine e- juice since they are the most safe Vaping techniques and also provides you extra high-end as well as fun while vaping. If you are able to vape without nicotine, you will certainly escape health issue which are caused by Pure nicotine along with found brand-new, pleasurable vaping methods.suorin

Bobby ley said...

hp envy 5010 setup
hp envy photo 6255 ìnstall
hp envy 5052 ìnstall

vanitha said...

Nice Blog!!!
Get the recent IEEE 2019 mini projects for learners in the Final Year of ECE , Cheap ECE-based mini projects for learners in Chennai

vanitha said...
This comment has been removed by the author.
madhusudhan creation said...

Thanks for Publishing the blog!!!
UV Gullas College of MedicineTraffic Analysis (UV Gullas College Traffic Analysis) is the technique by which emails are intercepted and examined to deduce information from communication patterns. Generally speaking, the more messages observed, intercepted and stored, the greater the deduction of traffic. Analysis of traffic UV Gullas of medical college may be carried out about military intelligence.

milo kith said...

hp envy 5012 setup
hp envy 7855 setup
hp envy 7858 setup

Prakash said...

More valuable post!!! Thanks for sharing this great post with us.
JAVA Training in Chennai
JAVA Course in Chennai
Java Training
Java classes in chennai
JAVA Training in Annanagar
java training in vadapalani
Digital Marketing Course in Chennai
Python Training in Chennai
Big data training in chennai
Selenium Training in Chennai

vanitha said...

Nice Blog!!!
Lyceum Northwestern University In producing leaders & academics, an ISO 9001:2008 acknowledged college with a wealthy heritage of 50 years.

Prakash said...

Great post!!! Thanks for sharing this wonderful blog with us...
SEO Training in Chennai
SEO Course in Chennai
SEO Training
seo classes in chennai
SEO training in Adyar
SEO training in Tnagar
Big data training in chennai
Hadoop training in chennai
Android training in Chennai
Selenium Training in Chennai

Sharmasandeep said...

Very good write-up. I certainly love this website. Thanks!

UI Development Training in Marathahalli

Full stack Development Training in Marthahalli Bangalore


UI Development Training in Bangalore


Angular Training in Bangalore

Sharmasandeep said...

There is definately a lot to know about this issue. I like all the points you've made.

Python Training in Marathahalli, Bangalore

Selenium Training in Marathahalli, Bangalore


Reactjs Training in Marathahalli, Bangalore

Sharmasandeep said...

Pretty! This has been a really wonderful post. Many thanks for providing this information.


Selenium Training in Marathahalli, Bangalore


Reactjs Training in Marathahalli, Bangalore

raj khan said...

Soma pill is very effective as a painkiller that helps us to get effective relief from pain. This cannot cure pain. Yet when it is taken with proper rest, it can offer you effective relief from pain.
This painkiller can offer you relief from any kind of pain. But Soma 350 mg is best in treating acute pain. Acute pain is a type of short-term pain which is sharp in nature. Buy Soma 350 mg online to get relief from your acute pain.

https://globalonlinepills.com/product/soma-350-mg/


Buy Soma 350 mg
Soma Pill
Buy Soma 350 mg online



Buy Soma 350 mg online
Soma Pill
Buy Soma 350 mg

dhanush kumar said...

This blog has very useful information about this topic which i am searching now, i gather some information from reading your blog
salesforce Training in Bangalore
uipath Training in Bangalore
blueprism Training in Bangalore

Punjabi song 2019 said...


New Punjabi Songs Download- Listen 2019 New Punjabi songs free online or Download Latest Punjabi Songs MP3.
Punjabi song 2019


charmidevan said...

valuable blog thanks for sharing it...waiting for next update...
Mobile Testing Training in Chennai
Mobile App Testing Training
Mobile Testing training in vadapalani
Mobile Testing training in Guindy
Mobile Testing training in Thiruvanmiyur
Manual Testing Training in Chennai
LoadRunner Training in Chennai
Photoshop Classes in Chennai
Spring Training in Chennai
QTP Training in Chennai

Customer Service Helpline Number USA said...

Office gives you the best involvement in your works. You can without much of a stretch do you work simple with the assistance of office application. It can gives you exceed expectations, word and PowerPoint so you can utilize it and makes your work simple.

Office.com/setup

Norton.com/setup

WWW.Office.com/setup

WWW.Norton.com/setup

HP Printer Expert said...

Good blog... Very valuable data is given on your blog.. Here is an approach to discover.
123.hp.com || 123.hp.com/setup || 123HP Setup || hp.com/setup || hp.com/123 || 123.hp.com setup || 123 HP Printer Setup || 123 HP Printer Support || 123 HP Setup and Install || 123hpcom || 123 HP Printer Install || 123hpcomsetup || 123 HP Wireless Setup || 123 HP Install || hpcom/123 || 123hpcominstall || 123HP Setup || 123 HP Smart App || Install 123 HP Printer || HP 123 Setup Scanner

«Oldest ‹Older   801 – 832 of 832   Newer› Newest»