7. Reinforcement Learning.pdf
0 likes116 views
The document discusses reinforcement learning techniques. It describes reinforcement learning as a method for solving interacting problems by considering past data to determine the next action. Reinforcement learning is also used in artificial intelligence to train machines through reward and punishment in tasks like walking. The document outlines reinforcement learning models including Upper Confidence Bound (UCB) and Thompson Sampling.
![UCB Python Code
# Importing the libraries
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
# Importing the dataset
dataset = pd.read_csv('Ads_CTR_Optimisation.csv')
# Implementing UCB
import math
N = 10000
d = 10
ads_selected = []
numbers_of_selections = [0] * d
sums_of_rewards = [0] * d
total_reward = 0
for n in range(0, N):
ad = 0
max_upper_bound = 0](https://image.slidesharecdn.com/7-221216061722-01daa25d/85/7-Reinforcement-Learning-pdf-39-320.jpg)
![UCB Python Code
for i in range(0, d):
if (numbers_of_selections[i] > 0):
average_reward = sums_of_rewards[i] / numbers_of_selections[i]
delta_i = math.sqrt(3/2 * math.log(n + 1) / numbers_of_selections[i])
upper_bound = average_reward + delta_i
else:
upper_bound = 1e400
if upper_bound > max_upper_bound:
max_upper_bound = upper_bound
ad = i
ads_selected.append(ad)
numbers_of_selections[ad] = numbers_of_selections[ad] + 1
reward = dataset.values[n, ad]
sums_of_rewards[ad] = sums_of_rewards[ad] + reward
total_reward = total_reward + reward
# Visualising the results
plt.hist(ads_selected)
plt.title('Histogram of ads selections')
plt.xlabel('Ads')
plt.ylabel('Number of times each ad was selected')
plt.show()](https://image.slidesharecdn.com/7-221216061722-01daa25d/85/7-Reinforcement-Learning-pdf-40-320.jpg)
7. Reinforcement Learning.pdf
- 2. Reinforcement Learning 2 Topics 7.1 Upper Confidence Bound 7.2 Thompson Sampling 7.3 Q-Learning
- 3. Reinforcement Learning • Reinforcement Learning is used to solve interacting problems where the data observed up to time t is considered to decide which action to take at time t + 1. • It is also used for Artificial Intelligence when training machines to perform tasks such as walking. Desired outcomes provide the AI with reward, undesired with punishment. Machines learn through trial and error. • Reinforcement Learning models: Upper Confidence Bound (UCB) Thompson Sampling
- 5. What is RL?
- 6. RL Process
- 7. RL Process
- 10. RL Concept- Reward Maximization
- 12. Upper Confidence Bound (UCB)
- 14. Difference between UCB & Thompson Sampling Algorithm 14
- 17. Reinforcement Learning •Reinforcement Learning is used to solve interacting problems where the data observed up to time t is considered to decide which action to take at time t + 1. •It is also used for Artificial Intelligence when training machines to perform tasks such as walking. Desired outcomes provide the AI with reward, undesired with punishment. Machines learn through trial and error. •Reinforcement Learning models: Upper Confidence Bound (UCB) Thompson Sampling https://www.amazon.com/gp/mpc/A18C1YWPOVXRKS
- 18. Multi-Armed Bandit Problem using Upper Bound confidence Algorithm
- 19. Upper Confidence Bound (UCB)
- 20. Upper Confidence Bound (UCB)
- 21. Upper Confidence Bound (UCB)
- 22. Upper Confidence Bound (UCB)
- 23. Upper Confidence Bound (UCB)
- 24. Upper Confidence Bound (UCB)
- 25. Upper Confidence Bound (UCB)
- 26. Upper Confidence Bound (UCB)
- 27. Upper Confidence Bound (UCB)
- 28. Upper Confidence Bound (UCB)
- 29. Upper Confidence Bound (UCB)
- 30. Upper Confidence Bound (UCB)
- 31. Upper Confidence Bound (UCB)
- 32. Upper Confidence Bound (UCB)
- 33. Eclat Output
- 34. Upper Confidence Bound (UCB)
- 35. Upper Confidence Bound (UCB)
- 36. Upper Confidence Bound (UCB)
- 37. Upper Confidence Bound (UCB)
- 38. Upper Confidence Bound (UCB)
- 39. UCB Python Code # Importing the libraries import numpy as np import matplotlib.pyplot as plt import pandas as pd # Importing the dataset dataset = pd.read_csv('Ads_CTR_Optimisation.csv') # Implementing UCB import math N = 10000 d = 10 ads_selected = [] numbers_of_selections = [0] * d sums_of_rewards = [0] * d total_reward = 0 for n in range(0, N): ad = 0 max_upper_bound = 0
- 40. UCB Python Code for i in range(0, d): if (numbers_of_selections[i] > 0): average_reward = sums_of_rewards[i] / numbers_of_selections[i] delta_i = math.sqrt(3/2 * math.log(n + 1) / numbers_of_selections[i]) upper_bound = average_reward + delta_i else: upper_bound = 1e400 if upper_bound > max_upper_bound: max_upper_bound = upper_bound ad = i ads_selected.append(ad) numbers_of_selections[ad] = numbers_of_selections[ad] + 1 reward = dataset.values[n, ad] sums_of_rewards[ad] = sums_of_rewards[ad] + reward total_reward = total_reward + reward # Visualising the results plt.hist(ads_selected) plt.title('Histogram of ads selections') plt.xlabel('Ads') plt.ylabel('Number of times each ad was selected') plt.show()
- 41. UCB Output
- 42. What is RL?
- 43. RL Process
- 44. RL Process
- 45. RL Definitions:
- 46. RL Definitions:
- 47. RL Concept- Reward Maximization
- 50. Markov Decision Process-Shortest Path Problem This is an example of EXPLOITATION