Machine Learning Methods for Data Analysis
In today's data-driven world, the ability to extract valuable insights from large datasets is a crucial skill. This course, "Machine Learning Methods for Data Analysis," is designed to equip you with the knowledge and practical skills to harness the power of machine learning techniques for effective data analysis.
Speakers: Javier Redondo, Qaisar Farooq, Wajid Ali
Gillespie Stochastic Algorithm (Wajid Ali, University of Liverpool, UK): In this lecture, we used the Gillespie Stochastic Algorithm (GSA) for generating synthetic data for a simple birth-death process and used least -square minimization method to estimate the parameters by fitting deterministic birth-death model to the data generated with GSA.
The Multi-Armed Bandit Problem (Javier Redondo Anton, University of Turin, Italy): Within machine learning, there are a wide variety of techniques that allow us to solve problems of very different types. In this lecture we will discuss the Multi-Armed Bandit problem, a problem in which a limited set of resources must be allocated among competing (alternative) choices in a way that maximizes their expected gain. This is a classic reinforcement learning problem that exemplifies the exploration-exploitation tradeoff dilemma. To solve it we will rely on several machine learning techniques, seeing in a simple way how this problem can be solved and the different variations that exist. In addition, we will see how these methods can be applied to domains as different as the choice of the best restaurant in town or the allocation of samples in a parameter sampling problem.
[16 December 2023] Markov Chain Monte Carlo (MCMC) Methods (Javier Redondo Anton, University of Turin, Italy): Markov Chain Monte Carlo (MCMC) methods encompass a set of algorithms designed for sampling from probability distributions. In this talk, we will look into the underlying principles, elucidating how and why these methods are effective across various fields, including biology and physics. Through simple examples, we will explore different implementations, shedding light on the mechanics of well-known algorithms such as the Metropolis-Hastings algorithm, Slice sampling, and Hamiltonian Monte Carlo.
Lecture#4:
Lecture#5: