Python Tutorial: What is Keras?

Key Takeaways

welcome to this course on deep learning I'm Mia and I'm very excited to be teaching you Karis here on betacam this course will have Kara's as a powerful tool to your arsenal Curren is a high-level deep learning framework to understand what is meant by that we can compare it to a lower level framework like TN building a neural network in theano can take many lines of code and it requires a deep understanding of how they work internally building and training this very same network in Karis only takes a few lines of code much quicker right girls is an open source deep learning library that enables fast experimentation with neural networks it runs on top of other frameworks like tential 30 on o or c NT k and it was created by Francie I researcher Francois Chawla while you scare us instead of slaughter low level libraries like tensorflow with Karis you can build industry ready models in no time with much less cool than Tia as we saw before and a higher extraction than that offer by tensorflow this allows for quickly and easily checking if a neural network will get your province off in addition you can do any architecture you can imagine from simple networks to more complex ones like our encoders convolutional or recurrent neural networks chaos models can also be deployed across a wide range of platforms like Android iOS where apps etc it is the best moment to be learning Charis Charis is now fully integrated into tension flow 2.0 so you can use the best of both worlds as needed and in the same code pipeline if as you dive deep into deep learning you find yourself needing to use low-level features for instance to have a final control of all your network applies gradients you could use tensor flow and tweak whatever you need now that you know better what Kara says and why to use it perhaps we shall discuss when and why to use neural networks in the first place neural networks are good feature structures since they've learned the best way to make sense of unstructured tape previously it was the domain Esper that had to set rules based on experimentation and heuristics to extract the relevant features of data neural networks can learn the best features under combination the camper from feature engineering themselves and that's why they are so useful but what are some structured data structured data is data that is not easily put into a table for instance sound video images etc it is also the type of data were performing feature engineering can be more challenging and that's why leaving this task to neural networks is a good idea if you are dealing we don't structured there you don't need to interpret the results and your problem can benefit from a known architecture then you probably should use neon networks for instance when classifying images of cats and dogs images are unstructured data we don't care as much about why the network knows it is a cat or a dog and we can benefit from a convolutional neural network so it is wise to use neural networks you will learn about the usefulness of convolutional neural networks later on in the course it is now time to review