Mediaspace scheduled maintenance: Aug 25, 2026 07:00 - 12:00 AM. During this time, videos will be temporarily unavailable. Check status updates.

Lecture

Policy Gradient Methods: Multiple Time Steps

In course

CS-456: Deep reinforcement learning

This course provides an overview and introduces modern methods for reinforcement learning (RL.) The course starts with the fundamentals of RL, such as Q-learning, and delves into commonly used approaches, like PPO and DQN. The course will introduce student ...

Description

This lecture covers Policy Gradient methods over multiple time steps, aiming to update policy parameters to maximize the total discounted reward. The slides present the derivation of these methods, including the calculation of accumulated rewards in episodes and the pseudo-code for the REINFORCE algorithm.

Instructors (2)

Wulfram Gerstner

Wulfram Gerstner is Director of the Laboratory of Computational Neuroscience LCN at the EPFL. His research in computational neuroscience concentrates on models of spiking neurons and spike-timing dependent plasticity, on the problem of neuronal coding in single neurons and populations, as well as on the link between biologically plausible learning rules and behavioral manifestations of learning. He teaches courses for Physicists, Computer Scientists, Mathematicians, and Life Scientists at the EPFL. After studies of Physics in Tübingen and at the Ludwig-Maximilians-University Munich (Master 1989), Wulfram Gerstner spent a year as a visiting researcher in Berkeley. He received his PhD in theoretical physics from the Technical University Munich in 1993 with a thesis on associative memory and dynamics in networks of spiking neurons. After short postdoctoral stays at Brandeis University and the Technical University of Munich, he joined the EPFL in 1996 as assistant professor. Promoted to Associate Professor with tenure in February 2001, he is since August 2006 a full professor with double appointment in the School of Computer and Communication Sciences and the School of Life Sciences. Wulfram Gerstner has been invited speaker at numerous international conferences and workshops. He has served on the editorial board of the Journal of Neuroscience, Network: Computation in Neural Systems', Journal of Computational Neuroscience', and `Science'.

Caglar Gulcehre

Official source

Related concepts (31)

Reinforcement learning

Reinforcement learning (RL) is an area of machine learning concerned with how intelligent agents ought to take actions in an environment in order to maximize the notion of cumulative reward. Reinforcement learning is one of three basic machine learning paradigms, alongside supervised learning and unsupervised learning. Reinforcement learning differs from supervised learning in not needing labelled input/output pairs to be presented, and in not needing sub-optimal actions to be explicitly corrected.

Derivative

In mathematics, the derivative shows the sensitivity of change of a function's output with respect to the input. Derivatives are a fundamental tool of calculus. For example, the derivative of the position of a moving object with respect to time is the object's velocity: this measures how quickly the position of the object changes when time advances. The derivative of a function of a single variable at a chosen input value, when it exists, is the slope of the tangent line to the graph of the function at that point.

Generalizations of the derivative

In mathematics, the derivative is a fundamental construction of differential calculus and admits many possible generalizations within the fields of mathematical analysis, combinatorics, algebra, geometry, etc. The Fréchet derivative defines the derivative for general normed vector spaces . Briefly, a function , an open subset of , is called Fréchet differentiable at if there exists a bounded linear operator such that Functions are defined as being differentiable in some open neighbourhood of , rather than at individual points, as not doing so tends to lead to many pathological counterexamples.

Second derivative

In calculus, the second derivative, or the second-order derivative, of a function f is the derivative of the derivative of f. Informally, the second derivative can be phrased as "the rate of change of the rate of change"; for example, the second derivative of the position of an object with respect to time is the instantaneous acceleration of the object, or the rate at which the velocity of the object is changing with respect to time. In Leibniz notation: where a is acceleration, v is velocity, t is time, x is position, and d is the instantaneous "delta" or change.

Derivative test

In calculus, a derivative test uses the derivatives of a function to locate the critical points of a function and determine whether each point is a local maximum, a local minimum, or a saddle point. Derivative tests can also give information about the concavity of a function. The usefulness of derivatives to find extrema is proved mathematically by Fermat's theorem of stationary points. The first-derivative test examines a function's monotonic properties (where the function is increasing or decreasing), focusing on a particular point in its domain.

Related lectures (56)

Tangent to Graph of a Function

Explores finding the equation of the tangent line to a function's graph at a point.

Derivatives: Definition and Properties

Explores the definition and properties of derivatives, including slopes of tangent lines and differentiability conditions.

Derivative Composition Theorem MOOC: Analysis I MOOC: Analysis I (part 5): Continuous and derivable functions, the derivative function

Covers the derivative of the composition of two functions and its applications.

Exact Linearization: Earth Dynamics and Stabilization

Explores exact linearization techniques for transforming non-linear systems into linear ones, emphasizing system stability.