文件列表:

agentzoo (0, 2015-12-07)
agentzoo\puckagent.json (36359, 2015-12-07)
agentzoo\wateragent.json (446607, 2015-12-07)
external (0, 2015-12-07)
external\.DS_Store (6148, 2015-12-07)
external\d3.min.js (150762, 2015-12-07)
external\highlight.pack.js (10183, 2015-12-07)
external\highlight_default.css (2642, 2015-12-07)
external\images (0, 2015-12-07)
external\images\ui-bg_diagonals-thick_18_b81900_40x40.png (418, 2015-12-07)
external\images\ui-bg_diagonals-thick_20_666666_40x40.png (312, 2015-12-07)
external\images\ui-bg_flat_10_000000_40x100.png (205, 2015-12-07)
external\images\ui-bg_glass_100_f6f6f6_1x400.png (262, 2015-12-07)
external\images\ui-bg_glass_100_fdf5ce_1x400.png (348, 2015-12-07)
external\images\ui-bg_glass_65_ffffff_1x400.png (207, 2015-12-07)
external\images\ui-bg_gloss-wave_35_f6a828_500x100.png (5815, 2015-12-07)
external\images\ui-bg_highlight-soft_100_eeeeee_1x100.png (278, 2015-12-07)
external\images\ui-bg_highlight-soft_75_ffe45c_1x100.png (328, 2015-12-07)
external\images\ui-icons_222222_256x240.png (6922, 2015-12-07)
external\images\ui-icons_228ef1_256x240.png (4549, 2015-12-07)
external\images\ui-icons_ef8c08_256x240.png (4549, 2015-12-07)
external\images\ui-icons_ffd27a_256x240.png (4549, 2015-12-07)
external\images\ui-icons_ffffff_256x240.png (6299, 2015-12-07)
external\jquery-1.11.2.min.js (95931, 2015-12-07)
external\jquery-2.1.3.min.js (84319, 2015-12-07)
external\jquery-ui.min.css (18195, 2015-12-07)
external\jquery-ui.min.js (100680, 2015-12-07)
external\jquery.flot.min.js (52966, 2015-12-07)
external\marked.js (28191, 2015-12-07)
external\mathjax.js (60573, 2015-12-07)
external\underscore-min.js (16523, 2015-12-07)
gridworld_dp.html (26862, 2015-12-07)
gridworld_td.html (34259, 2015-12-07)
img (0, 2015-12-07)
img\dpsolved.jpeg (49769, 2015-12-07)
img\lambda.png (11661, 2015-12-07)
img\policyiter.png (8654, 2015-12-07)
img\qsa.jpeg (44553, 2015-12-07)
... ...

# REINFORCEjs **REINFORCEjs** is a Reinforcement Learning library that implements several common RL algorithms, all with web demos. In particular, the library currently includes: - **Dynamic Programming** methods - (Tabular) **Temporal Difference Learning** (SARSA/Q-Learning) - **Deep Q-Learning** for Q-Learning with function approximation with Neural Networks - **Stochastic/Deterministic Policy Gradients** and Actor Critic architectures for dealing with continuous action spaces. (*very alpha, likely buggy or at the very least finicky and inconsistent*) See the [main webpage](http://cs.stanford.edu/people/karpathy/reinforcejs) for many more details, documentation and demos. # Code Sketch The library exports two global variables: `R`, and `RL`. The former contains various kinds of utilities for building expression graphs (e.g. LSTMs) and performing automatic backpropagation, and is a fork of my other project [recurrentjs](https://github.com/karpathy/recurrentjs). The `RL` object contains the current implementations: - `RL.DPAgent` for finite state/action spaces with environment dynamics - `RL.TDAgent` for finite state/action spaces - `RL.DQNAgent` for continuous state features but discrete actions A typical usage might look something like: ```javascript // create an environment object var env = {}; env.getNumStates = function() { return 8; } env.getMaxNumActions = function() { return 4; } // create the DQN agent var spec = { alpha: 0.01 } // see full options on DQN page agent = new RL.DQNAgent(env, spec); setInterval(function(){ // start the learning loop var action = agent.act(s); // s is an array of length 8 //... execute action in environment and get the reward agent.learn(reward); // the agent improves its Q,policy,model, etc. reward is a float }, 0); ``` The full documentation and demos are on the [main webpage](http://cs.stanford.edu/people/karpathy/reinforcejs). # License MIT.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：