For those who want to build their first Machine Learning model or who want to apply Machine Learning for practical purposes we offer two 90-minute workshops. Register, attend and learn.
INTRO TO MACHINE LEARNING AND SOME RESERVOIR COMPUTING
10:00 Saturday, April 8th • Workshop duration is 90 minutes.
The workshop will start with a broader introduction to Machine Learning (ML): what it is, its context, a bit of history, relation to AI and other fields, types of ML, basic concepts and challenges.
We will overview some of the simplest ML methods and basic classes of more advanced ones. More attention will be given to neural networks, deep learning, and lastly Reservoir Computing (RC) - a field to which Mantas has contributed and which has led to some of the most exotic ML setups, challenging the notion of computation.
We will conclude with a small working RC python demonstration of an echo state network performing a simple task.
The workshop will be beginner-friendly and hopefully will benefit all in getting a clearer general view, as well as some applicable specifics.
- Python (numpy, scipy, pyplot) - if you want to run the code yourself
Dr. Mantas Lukoševičius
Machine learning researcher at Kaunas University of Technology
Mantas has been doing Machine Learning since "before it was cool" around 2003. He has obtained master's and PhD in the field at Jacobs University Bremen, specializing in recurrent neural networks, did an internship at one of the world's top ML/Deep Learning labs at University of Montreal. He is currently back to Lithuania, doing ML research and applied data science projects as a scientist and freelancer.
Log file transformation into predictive ML model
12:00 Saturday, April 8th • Workshop duration is 90 minutes.
During workshop you will analyze a log file of online shop user click events. You will get to know the data and will solve a problem how to identify users which are most likely to purchase a product. You will learn how to transform log file into predictive features and create few ML model prototypes using popular ML algorithms such as XGBoost and random forests.
The workshop will be available both for R and python users. You will be provided with a working ML code and will be guided with step-by-step explanations. During workshop we will discuss basic principles of building and evaluating ML models and tips how to perform ML algorithm parameter tuning.
- R (data.table, dplyr, lubridate, xgboost)
- Python (numpy, pandas, scikit-learn, xgboost)
DATA SCIENTIST AT SCORIFY
Darius has over 5 years of experience in making machine learning models and has created over 20 models for financial and telco companies. During spare time he also participates in Kaggle (largest online data science competitions platform) and has achieved 5th overall rank among 50k+ online data scientists.
Let’s apply TeNsorFlow
14:00 Saturday, April 8th • Workshop duration is 90 minutes.
During workshop you will get brief introduction to tensorflow. You will learn what neural networks are and how to create them. You will design neural network which will learn to classify handwritten digits. You will get to know how to adapt pre-trained neural network to solve various image classification problems.
You will be provided with working neural network training code and it will be thoroughly explained.
During workshop we will discuss basic principles of neural network architecture design.
- Python (tensorflow, numpy, scikit-learn)
Algorithm Engineer at Neurotechnology
Andrius has spent last 6 years working on Machine Learning/DNN projects in Computer Vision. He has experience designing CNNs, using tensorflow and lasagne deep learning frameworks and scikit-learn library.