Machine Learning
Next Date:
in-house
in-house
Duration:
5 days
5 days
Language:
English, Hungarian
English, Hungarian
Tooling:
computers needed with python and anaconda installed
computers needed with python and anaconda installed
Exam:
optionally yes
optionally yes
Objective:
This training course teaches python developers the basic skills that are required for artificial intelligence projects. This course provides a broad introduction to machine learning, supervised learning, unsupervised learning and reinforcement learning.
During their five days in the classroom students review the basic theory of machine learning, regression models and classification problems. The focus will be on Neural Networks and Deep Learning, and frameworks with which a novice learner can also train and run algorithms on their own laptop during the course. After the course the participant should be able to learn further machine learning technics on his/her own (with Tensorflow and Keras), and he or she will be able to apply these techniques in the real life projects within the company.
The course is available in the following programming languages: Python.
Participants:
Software developers with minimum 1 year Python background or years of practice in other programming languages. (The course does not require expert python knowledge)
Some mathematical background and basic understanding of matrices helps.
Content:
This course is mainly constructed for programmers who need to learn how to build Machine Learning applied applications and/or who want to get introduction to the topic. The course has a lot of examples and hands-on excercises
Modules:
Basic Theory:
- Definition of AI
- Definition of ML
- Different Methods of ML
- Logistic Regression
- Neural Networks
- Hyper Parameters
- SVM
- Unsupervised Learning
- Anomaly Detection
Deep diving (with hands-on examples):
- Brief Intro to Tensorflow and Keras
- Simple Logistic Regression examples
- Naive Bayes
- Shallow Neural Networks
- Deep Neural Networks
- Convolutional Networks
- Recurrent Neural Networks
- Reinforcement Learning
- Improving Neural Networks
- Hyper Parameter Tuning
- Regularization
- Optimization
- Pitfalls to avoid
Misc:
- Classroom with computers is required
- English handout and the source code of the demo applications will be provided for the participants.
- 2 whiteboards and a beamer with HDMI is required.