On 23 April, the second webinar in the cycle Fundamentals of Machine Learning, Supervised learning for regression and classification learning tasks, will take place at Rīga Stradiņš University (RSU).
Supervised learning for regression and classification learning tasks. Supervised learning becomes regression learning task in scenario where the dependent variable is a continuous value. Linear regression will be discussed, prediction error calculation and techniques for minimization of prediction error.
The learning task is called classification if the dependent variable is categorical. Multiple algorithms will be discussed - Logistic Regression, Naïve Bayes, K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Decision Trees and Random Forests. Algorithm performance calculation will be covered so you can compare performance of different algorithms applied to the same training and test data sets. Performance calculation will be based on analytics of confusion matrix.
Unsupervised learning to look for previously undetected patterns in a data set with no pre-existing labels. Two of the main methods used in unsupervised learning are cluster analysis and principal component analysis.
Principles of hyperparameters and grid search will be covered.
In the practical part you will build machine learning pipeline in Python to create a solution for solving regression or classification task depending on selected problem.
About the instructor
Uldis Doniņš is the Head of the Information Systems Unit of the IT Department at RSU. He holds a PhD (Dr.sc.ing.) in Computer Science and his field of study is software modeling and modeling formalisation. Uldis has expanded his knowledge and experience in the fields of machine learning and data intensive computing at the University at Buffalo (State University of New York, USA), School of Engineering and Applied Sciences. Being a part of Artificial Intelligence Machine Learning provides computer learning and decision-making based on the provided data that can be developed using supervised, unsupervised or reinforcement learning models. Data intensive computing deals with diverse data formats, storage models, application architectures, programming models and algorithms and tools for large-scale data analytics.
- About the webinar cycle
As the power and capabilities of computing increases, Artificial Intelligence solutions takes a greater role to perform and execute various processes. Being a part of Artificial Intelligence, Machine Learning provides computer learning and decision-making based on the provided data. Seminar is intended to provide insight into Machine Learning and its algorithms covering supervised and unsupervised learning, including data processing and application for machine learning solutions. Participants will get hands-on experience in implementing machine learning solutions by using Python which currently is one of the most popular programming languages.
The practical part is based on individual work on implementing machine learning project by using Python.
Next webinar in this series
|21 May||Machine learning project presentations by each participant and discussions|