Course Information
Registration period
Year-round Recruitment
Course Level
Study Mode
Duration
10 Hour(s) 4 Minute(s)
Language
English
Taught by
Francisco Juretig
Rating
Course Overview
Build complex deep learning algorithms easily in Python
Do you want to build complex deep learning models in Keras? Do you want to use neural networks for classifying images, predicting prices, and classifying samples in several categories?
Keras is the most powerful library for building neural networks models in Python. In this course we review the central techniques in Keras, with many real life examples. We focus on the practical computational implementations, and we avoid using any math.
The student is required to be familiar with Python, and machine learning; Some general knowledge on statistics and probability is recommended, but not strictly necessary.
Among the many examples presented here, we use neural networks to tag images belonging to the River Thames, or the street; to classify edible and poisonous mushrooms, to predict the sales of several video games for multiple regions, to identify bolts and nuts in images, etc.
We use most of our examples on Windows, but we show how to set up an AWS machine, and run our examples there. In terms of the course curriculum, we cover most of what Keras can actually do: such as the Sequential model, the model API, Convolutional neural nets, LSTM nets, etc. We also show how to actually bypass Keras, and build the models directly in Theano/Tensorflow syntax (although this is quite complex!)
After taking this course, you should feel comfortable building neural nets for time sequences, images classification, pure classification and/or regression. All the lectures here can be downloaded and come with the corresponding material.
Course Content
- 7 section(s)
- 39 lecture(s)
- Section 1 Introduction
- Section 2 Keras fundamentals
- Section 3 Scikit-learn and Keras
- Section 4 Classes for images
- Section 5 Multilayer Perceptron
- Section 6 Convolutional Neural Nets
- Section 7 Recurrent neural networks
What You’ll Learn
- Use Keras for classification and regression in typical data science problems, Use Keras for image classification, Define Convolutional neural networks, Train LSTM models for sequences, Process the data in order to achieve to the specific shape that Keras expects for each problem, Code neural networks directly in Theano using tensor multiplications, Understand what are the different layers that we have in Keras, Design neural networks that mitigate the effect of overfitting using specific layers, Understand how backpropagation and stochastic gradient descent work
Skills covered in this course
Reviews
-
MMaaz Khan
The Lectures are sometimes abruptly ending. It would be good if there is a flow explained first and lectures are ended on a proper note. Otherwise the concepts were explained well.
-
JJessie Lin
I wish I could understand the math behind convolutional layer and backpropegation a bit better.
-
RRemco Jansen
Good, but somewhat messy, not always explaining the important parts, like why a choice is made and using a mouse as a 'pencil'. But the information you get is quite good. You get that he knows the stuff, but he can't not always articulate it well.
-
AAshish Kumar
The course is definitely useful to get grips over CNN with good examples along with their code made available. I could clearly understand the reasons behind using different layers, parameters and so on. Takeaway for me is that this course has equipped me with the basic knowledge that I can use to start exploring new datasets and try solving them, try more complex problems to get better understanding. Following are few things that could have made this course even better: 1. The provided examples in the course are definitely a must and could not have been avoided. Jumping directly to complex problems is a strict No and the author made sure that he did not commit this mistake. However along with the provided example problems, if the the author had added few problems that are more complex, then that would have been really helpful for the students of this course. For example the given image sets under CNN are pretty easy for the network to solve, there are many easy differences in the images that the network will find it easier to solve, like nuts and bolts, riverside, etc. If some complex image sets could have been added that looked pretty similar and yet the model was able to classify, that would have been great. 2. Instead of using mouse to write and explain, a stylus could have been used. 3. I have subscribed to other similar courses and in almost all courses, I find one thing missing which is missing in this course as well is the 'comprehensiveness'. Think from the student's point of view, at the best they may posses a CPU enabled system on which they might be learning all these. Imagine when they try to test different data inputs after training a model, they have to train the model everytime they log-off and login their PC, its annoying. The authors can teach on how to save the weights and reload them the next time the student want to use the model. Surprisingly I have not seen any course doing this. Although this does not reduce the usefulness of this course but having it helps. I am fully satisfied with this course. Expecting similar great courses from him in future. Any course from the author on some Advanced CNN or Unsupervised learning or Real world problem solving using Deep Learning, etc will be great(provided its on windows and not ubuntu or mac), waiting for that. Thanks.
