Practical tasks
At each sesion you will have a task to work on your own. We work at Google Collaboratory.
Practical tasks
The basis of the course will be your work with ready-made code for research and better understanding of neural networks or writing your own code.
- 1Weekly assignmentEach topic is consolidated by completing a practical task. At the beginning of the course, the tasks will be simple, most often - to understand the code and test different variants of the ready neural network. In the second half of the course, the tasks will already include independent writing of the code and conducting complex research
- 2Curators will check your assignmentsAll assignments are checked by your curators. You will get the necessary feedback - on the written code and on what conclusions you have drawn. They can also give additional assignments to advanced participants.
- 3Competitions at KaggleSome of the training exercises we do on Kaggle, it is very useful to see how well you have mastered neural networks compared to other course participants.
Take a look at examples of our classes
"Basics of neural network programming on Python"
In this webinar you will learn first about the whole area of neural networks and artificial intelligence and in the second half - the basics of neural networks and writing an example of neural networks for handwritten digit recognition
"Network training quality evaluation. Overfitting. Training, training and test samples"
In this webinar you will learn one of the key themes of neural networks - working with the training and test sample, as well as continue working with convlutional networks and image recognition
"Convolutional neural networks. Classification of objects in images"
In this webinar, you'll learn about convolutional neural networks. This is the most powerful tool for working with any image, their principle of operation is built similar to the work of the human visual system. We'll analyze convolutional networks on the example of recognition of various objects
All work is done
at Google Collaboratory
at Google Collaboratory
Google's servers enable high-quality neural network training, including the ability to provide GPUs. Due to this, there are no requirements for the power of your computer and software, all you need is the browser Google Chrome.
Competitions at Kaggle
Kaggle is a website dedicated to topics related to machine learning and neural networks. Employers often check the Kaggle profile.
In the Kaggle profile, you can see all the tasks that you have completed. During training, the profile will become your portfolio. You will see your place in group and world rankings. Everything is simple and transparent.
In the Kaggle profile, you can see all the tasks that you have completed. During training, the profile will become your portfolio. You will see your place in group and world rankings. Everything is simple and transparent.
This may be of interest
Other pages of the website