Deploying Artificial Intelligence Applications
Francesco Pugliese
DATE Wednesday, 20th of March 2019
LOCATION Hub di LVenture Group e LUISS EnLabs (via Marsala 29/H, Rome)
This Workshop will lead the attendees towards the most exciting breakthroughs in the modern Artificial Intelligence by implementing code in Python and Keras. The participant will learn how to implement AI programs very easily and effectively by putting hands on Python constructs and Keras Core.
Combo Special offer! Save more than 22% on workshop + Laggard conference ticket full price (€ 330)! Only 20 tickets available.
Click here to know how to obtain these discounts.
LANGUAGE
Italian
LEVEL
Begginer
DURATION
The workshop is full-day (8 hours) from 09:00 to 18:00, with one hour lunch break.
CHECK IN 8:30 - 09:00
PRICES
114 € the first 20 tickets;
190 € until the 19th of March;
Combo Special offer! Save more than 22% on workshop + Laggard conference ticket full price (€ 330)! Only 20 tickets available.
Click here to know how to obtain these discounts.
Francesco Pugliese is an Istat’s Researcher and co-founder of Deep Learning Italia. He has a BsC and MsC in Computer Science, at the University of Rome “La Sapienza” and a PhD in Cognitive Science.
ABSTRACT
This Workshop will lead the attendees towards the most exciting breakthroughs in the modern Artificial Intelligence by implementing code in Python and Keras. The participant will learn how to implement AI programs very easily and effectively by putting hands on Python constructs and Keras classes. We will tackle the 3 most appealing case of study in AI nowadays: CV, NLP and Chatbots. This way, developers will get able to write cutting-edge Keras back-end code by using packages of the Keras Core such as Keras.Layers and useful libraries like OpenCV, Scikit Image, Scikit Learn, Numpy, Pandas
TABLE OF CONTENTS
Theory Module (4 hours)
– Introduction to Artificial Intelligence in Research, Business, and every-day-life.
– Explanation of all Artificial Intelligence Models with a particular focus on Artificial Neural Networks.
– Deep Neural Networks: Convolutional Neural Networks, Long Short Term Memories, Sequence to Sequence Models.
– Pre-processing: Word Embeddings (Skip-Gram, CBOW, etc.), Open CV for Images preprocessing.
– Model Regularizers: Dropout, BatchNormalization, Data Augmentation, L1/L2, etc.
– Training Algorithms and Regularizers: SGD, Adam, K-fold Cross Validation, Grid-Search, etc.
Practice Module (4 hours)
– Case Study 1: Building a complete Computer Vision (CV) System in Python/Keras. This section will explore how to implement “hands-on” advanced Convolutional Neural Networks (lenet, resnet, etc.), data augmentation, dropout, batch normalization, image preprocessing, ensemble models, etc.
– Case Study 2: Building a complete Natural Language Processing (NLP) System in Python/Keras. This section will explore how to implement “hands-on” a language model with word2vec, 1-Dimensional convolutional neural networks for NLP, text mining and preprocessing for deep learning, etc.
– Case Study 3: Building a complete Neural Chatbot in Python/Keras. This section will explore how to implement “hands-on” an advanced chatbot by seq2seq models, dynamic memory networks, etc.
TRAINING OBJECTIVES
Understanding Artificial Intelligence programming principles
Understanding Deep Learning theory and programming
Learning Convolutional Neural Networks
Long Short Term Memories
Residual Networks
Generative Adversarial Networks and other complex Deep Neural Networks
Learning how to program Python AI programs with Keras
Implementing complex AI models in keras such as NLP and Images classfiers
Text Summarization and Chatbots
Measuring by metrics accuracy of models
Implement cutting-edge “businessable” AI applications
WHO THE WORKSHOP IS DEDICATED TO?
People who would like to learn fundamentals of modern AI, on how to quickly program python code for AI applications, people interested in research on AI, people interested in building a Business and/or a StartUp around AI.
PREREQUIREMENTS
No experience in Artificial Intelligence programming.
Basic knowledge of programming in Python is recommended, but not strictly necessary.
General knowledge of mathematical analysis, linear algebra and statistics should be due.
HARDWARE AND SOFTWARE REQUIREMENTS
Attendees should have a laptop equipped with a GPU possibly but not necessarily. Laptop can have CPU only, but computations risk to be more time-consuming thand without GPU. If possible, attendees should have Python installed with Anaconda distribution, TensorFlow and Keras.
Deep Learning Italia offers tutorials for installing all the SW stack needed for our classes at the following links:
Italiano: http://www.deeplearningitalia.com/installazione-di-kerastensorflow-theano-su-windows/
English: http://www.deeplearningitalia.com/installation/ A virtual machine not supporting GPU with Keras installed will be provided to very beginners.
WARNING
Seats are limited.
The workshop will be held only if the minimun number of attendees is reached.
Combo Special offer! Save more than 22% on workshop + Laggard conference ticket full price (€ 330)! Only 20 tickets available.
Click here to know how to obtain these discounts.
