MATLAB House :: Channel
364 subscribers
13 photos
72 videos
2 files
48 links
— این کانال جهت تبادل هر چه بهتر اطلاعات و اشتراک دانش در حوزه نرم‌افزار #متلب ایجاد شده است.
— گپ:@MATLABHOUSE
— آموزش‌ها و پروژه‌های تکمیلی در justeducation.ir قرار خواهد گرفت.
Download Telegram
Comparative Analysis of Bandit Algorithms for Optimal Decision-Making

Explore
a comprehensive comparative analysis of various bandit algorithms used in reinforcement learning for optimal decision-making. This video showcases the implementation and evaluation of different methods such as Greedy, Epsilon-Greedy, UCB, and more, highlighting their strengths and performance in selecting optimal actions. Gain insights into the trade-off between exploration and exploitation strategies and learn how these algorithms can enhance decision-making systems. Join us for a deep dive into the world of bandit algorithms and their applications.

YouTube: https://youtu.be/K2dPVza-pSQ


🆔 @MATLAB_House

@MATLABHOUSE

#ReinforcementLearning #BanditAlgorithms #DecisionMaking #ExplorationVsExploitation #OptimalActionSelection #MachineLearning #DataScience #AI #CodeImplementation #AlgorithmComparison #PerformanceAnalysis
Reinforcement Learning in Gridworld: Solving the Windy Grid Problem

Watch this video showcasing the implementation of a reinforcement learning algorithm in solving the Windy Grid Problem. The algorithm uses Q-learning with epsilon-greedy exploration to navigate a gridworld with varying wind powers. Learn how the agent learns to reach the goal by optimizing its actions based on rewards and Q-values. The video includes visualizations of the grid, wind powers, and the agent's path.

YouTube: https://youtu.be/AiI_4flFmYc

🆔 @MATLAB_House

@MATLABHOUSE

#ReinforcementLearning #Qlearning #Gridworld #WindyGridProblem #ArtificialIntelligence #MachineLearning #CodingTutorial #Python #Algorithm #AI
Media is too big
VIEW IN TELEGRAM
✳️Inverted Pendulum Control: RL + MPC Implementation

1️⃣ Reinforcement Learning Features:
- Q-Learning for system identification
- Self-learning pendulum balancing
- No prior model needed

2️⃣ MPC Implementation:
- Real-time optimization
- Constraint handling
- Precise position/angle control

3️⃣ Hardware:
- DC motor (50:1 gearbox)
- Dual encoders
- STM32 controller
- Custom PWM driver

4️⃣ Performance:
- Upright stabilization
- Disturbance rejection
- Accurate tracking

By: Javad Safaei
Supervisor: Naser Pakar
🔹Telegram:
🆔 @MATLAB_House
@MATLABHOUSE

#InvertedPendulum #ReinforcementLearning #ModelPredictiveControl #Robotics #ControlSystems #Engineering
1