The following environment is recommended:
Python 3.9 , numpy 1.21.5, pandas 1.4.2, Spyder 5.1.5


***For Synthetic dataset empirical studies,

Run "k_10_d_2.py" to see the regret plot of different algorithms when K=10 and d=2;
Run "k_100_d_2.py" to see the regret plot of different algorithms when K=100 and d=2;
Run "k_500_d_2.py" to see the regret plot of different algorithms when K=500 and d=2;
Run "k_10_d_20.py" to see the regret plot of different algorithms when K=10 and d=20;
Run "k_10_d_50.py" to see the regret plot of different algorithms when K=10 and d=50;


***To see the Click-Through Rate (CTR) plot for top-20 movies,
First make sure the code in Line 42 "movielens_preprocessing.py"
is "actions = data.groupby('movie_id').size().sort_values(ascending=False)[:20]", 
if not please modify it.
Second Run "movielens_preprocessing.py".
Lastly Run "Movielens_K_=20.py" and then you can see the CTR plot.



***For MovieLens dataset empirical studies,

***To see the Click-Through Rate (CTR) plot for top-50 movies,
First make sure the code in Line 42 of "movielens_preprocessing.py" is 
"actions = data.groupby('movie_id').size().sort_values(ascending=False)[:50]", 
if not please modify it.
Second Run "movielens_preprocessing.py".
Lastly Run "Movielens_K=50.py" and then you can see the CTR plot.

***To see the Click-Through Rate (CTR) plot for top-100 movies,
First make sure the code in Line 42 "movielens_preprocessing.py"
is "actions = data.groupby('movie_id').size().sort_values(ascending=False)[:100]", 
if not please modify it.
Second Run "movielens_preprocessing.py".
Lastly Run "Movielens_K=100.py" and then you can see the CTR plot.


The "movielens_preprocessing.py" is referenced by Computational Learning Lab, Dept. of Computer Science and Information Engineering, National Taiwan University "https://github.com/ntucllab/striatum/blob/master/examples/movielens_preprocess.py". Once the paper is accepted, we will add this reference in our paper.


 