One-Week Hands-On Training
in Deep Neural Networks

Eventbrite - Deep Learning: One-Week Hands-On Training in Deep Neural Networks

Deep Learning program will give you practical skills to work with Deep Neural Networks and help in overcoming the entry threshold to the new field. The program includes three main tracks: Image Processing, Text Processing, and Business Applications.

During the program week each participant will work on an own cloud GPU-server to complete lab work.

Software used: Caffe, Keras.

After the program you will be able to:

  • Work on real-life projects that require Deep Learning,
  • Build applications with image and text processing under the hood,
  • Participate in Kaggle competitions.

The program assumes that participants already have some background in Machine Learning and Data Science, basic skills in Python and Linux.

Class schedule: 2 full Sat from 10am to 6pm, and 3 weeknights Mon, Wed and Fri - 6pm to 9pm.


June 24 - July 1, 2017

Class Schedule

Sat 10am - 6pm
Mon, Wed, Fri 6pm - 9pm


On Campus Classes
Startup Temple
910 Broadway, San Francisco


$ 2,500



Short Syllabus

  • Images

    • • Intro to Deep Learning
    • • Neural Networks basics
    • • Convolutional Neural Networks (CNN)
    • • Case studies: LeNet, AlexNet, …
    • • Practice with Caffe: developing an image classificator
    • • Transfer learning, detection, and segmentation
    • • Style transfer
    • • Super-resolution images
  • Texts

    • • Recurrent Neural Networks (RNN)
    • • Text classification using RNNs, CNNs
    • • Practice with Keras/Theano: sentiment analysis of texts
    • • Sequence learning and seq2seq: translation, chatbots, speech recognition
    • • Case studies: chatbots
    • • Image captions: CNN and RNN
  • Business

    • • Business applications of Deep Learning and Artificial Intelligence
    • • Key business metrics for DL and AI projects

Deep Learning Alumni

  • In my opinion, the program is beautifully balanced between theory and practice. I had no idea that it was possible to pack such amount of material in such a short time. As a result, I got a clear picture from the history of the field to the state of the art techniques. Now I know where to start my own projects and where to go next. After the individual work during the week of the training I got the feeling that it wasn’t so hard to build a deep neural network for basic image classification. Also it was very convenient that we got our own GPU-servers with the pre-built software. I’m really happy that in such short time I got what I needed. I will definitely recommend this program to my friends and colleagues.
    Igor Parfenov
    co-founder of ML-consulting company,
    ex-CTO of several software engineering companies
  • It was my second educational program at New Professions Lab. The first was Big Data Specialist. The principles are the same - a lot of practice in a short time in a group of great people. In one week the complicated things were explained in simple words (not vice versa) and we were given a competent methodology of working with deep learning algorithms. Thanks a lot!
    Kirill Danilyuk
    co-founder of ML-consulting company,
    ex-product manager at Yandex
    and data products lead at ZeptoLab (famous by mobile game “Cut the Rope”)
  • It was a really useful program for me. Literally, during the first day I overcame the entry threshold because I’d had the image of deep learning as a very complicated and a hard thing before. The information about the current state of the industry, algorithms, principles gave the basis for my future self-education in the field. The practical exercises helped to understand how to deal with real-life tasks. Although we discussed many things during the program, I regretted that it finally ended. I’d like to continue my education maybe in a future longer program from New Professions Lab.
    Oleg Khomyuk
    Lead Data Scientist at ConsultantPlus,
    ex-project manager at Yandex