### Description:
#### 1. cold_start_analysis:
Analyses the performance of different approaches in case of a new user or a user with less number of interaction with the system, namely the cold start problem. Computed the rmse and mae for those customers who have rated less than 18 books and also who have rated more than 1000 movies. <br>
For less interactions, content based and item-item based collaborative filtering approaches work better. As the number of interactions per customer increases, SVD and collaborative approaches work better.
#### 2. combined_model:
Combination of different surprise model results by applying weighted linear combination to generate final rating.
#### 3. content_based_recommendation:
Genreating user and movie vectors based on genre and predicting the ratings for movies in test data.
#### 4. evaluating_recs:
Code for Precision, Recall, F-1 score and NDCG.
#### 5. generating_predictions:
Generating rating predictions for test data using surprise library.
#### 6. hybrid_model:
Code for the hybrid model based on combining recommendations from different models such as content based, CF, SVD to improve accuracy and quality of recommendations.
#### 7. knn_analysis:
Analysis of KNN algorithms by changing different parameters like:
* number of neighbors
* similarity metrices
* user v/s item based CF
#### 8. model_hyperparameter_tuning:
Fine-tuned surprise models by experimenting with different hyperparameters for training and model. Compared models based on RMSE and MAE.
#### 9. movie_era_based_recs:
Content based approach to include the time period in which the movie was launced in the user vector. This method personalizes the users recommendations to include this feature.
#### 10. movie_similarity_based_recs:
Content based approach to include the user's genre preference and recommend movies similar to user's highly rated movies.
#### 11. movie_year_analysis:
Experiments with the year of the movie release. Analysed the distribution of data and determine the appropriate era intervals to classify movies. Used the content based approach to form a user vector based on the era preference.
#### 12. popularity_model:
Model which uses the popularity attribute as well as the average rating and voter count in the TMDB data to generate popular movies genre wise. The genres are determined using the IMDB data.
#### 13. preprocessing:
Code for spliting the data into training and testing set for each user such that 80% ratings are in training and 20% are for testing.
#### 14. surprise_model_predictions:
Code for generating ratings for test data using surprise models such as KNN (CF), SVD, Baseline approach, Slopeone etc.
#### 15. surprise_model_recs:
Comparison between the surprise models based on test data ratings (RMSE and MAE) and quality of recommendations (precision, recall, ndcg, f-measure).
#### 16. test_ndcg:
Code to test implementation of [NDCG metric](https://en.wikipedia.org/wiki/Discounted_cumulative_gain) for evaluting recommendations.
没有合适的资源?快使用搜索试试~ 我知道了~
使用基于内容、基于协同过滤、基于SVD和基于流行度的方法设计了一个电影推荐系统_Jupyter Notebook.zip
共36个文件
ipynb:13个
png:11个
py:3个
1.该资源内容由用户上传,如若侵权请联系客服进行举报
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
版权申诉
0 下载量 96 浏览量
2023-04-23
10:46:51
上传
评论
收藏 5.15MB ZIP 举报
温馨提示
使用基于内容、基于协同过滤、基于SVD和基于流行度的方法设计了一个电影推荐系统_Jupyter Notebook.zip
资源推荐
资源详情
资源评论
收起资源包目录
使用基于内容、基于协同过滤、基于SVD和基于流行度的方法设计了一个电影推荐系统。_Jupyter Notebook_Python.zip (36个子文件)
Movie-Recommendation-System-master
.DS_Store 10KB
Results
.DS_Store 8KB
Final_model_results.xlsx 7KB
algo_results.csv 2KB
images
KNN_similarity.png 35KB
genre_based_popularity.png 118KB
ndcg.png 115KB
Algo_analysis.png 95KB
vector_generation.png 70KB
Hybrid_Model.png 52KB
rating.png 6KB
mae_rmse_including_pearson.png 161KB
prec_recall_fm.png 120KB
genre_distribution.png 177KB
knn_neighbors.png 55KB
README.md 259B
Code
popularity_model.ipynb 75KB
evaluating_recs.py 5KB
content_based_recommendation.ipynb 60KB
knn_analysis.ipynb 164KB
surprise_model_predictions.ipynb 22KB
movie_similarity_based_recs.ipynb 60KB
combined_model.ipynb 10KB
preprocessing.ipynb 43KB
test_ndcg.py 3KB
movie_year_analysis.ipynb 118KB
model_hyperparameter_tuning.ipynb 548KB
hybrid_model.ipynb 49KB
movie_era_based_recs.ipynb 35KB
README.md 3KB
generating_predictions.py 5KB
cold_start_analysis.ipynb 28KB
surprise_model_recs.ipynb 76KB
Report.pdf 1.48MB
README.md 1KB
Presentation.pptx 2.29MB
共 36 条
- 1
资源评论
快撑死的鱼
- 粉丝: 1w+
- 资源: 9156
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功