Introduction to Machine Learning with Scikit-learn - your first pipeline (post-conference workshop)
ABOUT WORKSHOP
During this workshop, we will focus on creating end-to-end Machine Learning pipeline for binary classification, including data preprocessing, model selection and results evaluation.
The workshop is dedicated to developers with knowledge of Python, who want to get familiar with Machine Learning concepts.
Outcomes of workshop:
- Intuition about main algorithms currently used in the industry
- Understanding how to create end-to-end Machine Learning solutions
- Essential chunks of code for your further ML projects
The workshop will be conducted in Polish.
REQUIRED SKILLS
- Python (numpy, pandas, matplotlib)
- Maths - basic knowledge of calculus and algebra
- Statistics - core concepts like distribution, mean, variance, correlation
AGENDA
1. Introduction, naming conventions, data sources
- Data visualization
- Missing values
- Outliers
- Categorical Variables
- Feature Scaling
- Splitting dataset into train, valid and test chunks
3. Classification problem:
- Problem specification
- Evaluation methodology
4. Algorithms:
- Logistic Regression
- K-Nearest Neighbors
- Support Vector Machines (+ kernels)
- Naive Bayes
- Decision Trees
- Random Forest
5. Best practices and training strategy
MINIMUM REQUIREMENTS
This is a BYOD (bring your own device) workshop, so remember to bring your own laptop.
The workshop will run for 4 hours from 9 AM until 1PM. There will be a few coffee breaks.
TRAINER - WALDEMAR KOŁODZIEJCZYK
Artificial Intelligence Developer z doświadczeniem przy budowaniu modeli operujących na szeregach czasowych. Interesuje się algorytmiką, rozwiązaniami uczenia ze wzmocnieniem oraz szeroko rozumianym statystycznym przetwarzaniem danych.
WORKSHOP FEE INCLUDES
- Half-day workshop
- Coffee & Tea
- Wi-Fi access
- Entrance ticket for Confitura conference (Sat, 29 June 2019)