Erdi güngör bbs
- 3. Content Define Objective Prepare Data Algorithms Interpret Results Data preparation Feature creation Classification Predictive models 3 Business objective
- 4. Objective The online shopping mall problem To make easy to fill basket with personal favorites of the customers What will be in the next market basket ?
- 5. Objective Supervised Learning Unsupervised Learning (Apriori, Collaborative Filtering) Deep LearningApproach MarkovChain
- 6. Content Define Objective Prepare Data Algorithms Interpret Results Data preparation Feature creation Classification Predictive models 6 Business objective
- 7. Prepare Data / Source Data Open source data Aisle – 135 record Department – 21 record Order Products Prior - 1048576 Order Products Train - 1048576 Orders- 3421083 Products - 49689
- 8. Prepare Data / Base Data Data Preparation Train set: 13863746 -23 Test set: 4833292 -24 Order_eval _set User Order Order_2 Order_2_ eval_set Products _2 Reordered Test 1234 4 1 Prior A 0 Test 1234 4 1 Prior B 0 Test 1234 4 1 Prior C 0 Test 1234 4 2 Prior D 0 Test 1234 4 2 Prior B 1 Test 1234 4 2 Prior E 0 Test 1234 4 3 Prior F 0 Test 1234 4 3 Prior G 1 Test 1234 4 3 Prior E 1 Order_ev al_set User Order Products Reordere d Prior 5678 10 X 0 Prior 5678 10 Y 0 Prior 5678 10 Z 0 Prior 5678 11 W 0 Prior 5678 11 Q 0 Prior 5678 11 Z 1 Train 5678 12 W 1 Train 5678 12 Q 1 Train 5678 12 Z 1
- 9. Prepare Data / Features Customer Based • Total number of orders / products / unique products / unique department / unique aisle / unique order day • Sum / Avg. of reordered products • The customer period • Avg.Time between orders • Rate of reordered products in total products • The order count after related product Customer – Product Based • Avg. sequence in market basket • Sum of reordered information of product • The max./avg. hour to order product • The min. / max. order number of product • The most popular hour for product Clustering Label • Kmeans 5 cluster Wrt percentage usage of sub-segment in over all
- 10. Content Define Objective Prepare Data Algorithms Interpret Results Data preparation Feature creation Classification Predictive models 10 Business objective
- 11. Algorithms DecisionTrees, Naive Bayes, Instance Based, Logistic Regression, SupportVector Machine, Regression Ensemble Learning: Random Forest, AdaBoost, Gradient Boosting Neural Networks
- 12. Algorithms Train-Test Split ( K-crossValidation) Extreme Gradient Boosting (XGBoost)* Random Forest Logistic Regression DecisionTree Gradient Boosting
- 13. Algorithms Extreme Gradient Boosting (XGBoost) Implementation of gradient boosted decision trees Regularization Parallel processing High flexibility ( all types of data) Handling missing values and tree pruning
- 14. Algorithms Extreme Gradient Boosting (XGBoost) Parameters Objective (Reg:Logistic): Learning objective Eval_metric (Logloss): Evaluation metric/Negative log-likelihood Eta (0.1): Shrinkage the feature weights to prevent overfitting Max_depth (7): Maximum depth of tree Min_child_weight (9) :Minimum sum of instance weight Gamma (0.80): Minimum loss reduction to make further partition Subsample (0.76): Subsample ratio of the training instance Colsample_bytree (0.95): Subsample ratio of columns when constructing each tree Alpha (0) : L1 regularization term on weight Lambda (1) : L2 regularization term on weight
- 15. Algorithms Extreme Gradient Boosting (XGBoost) Dmatrix : data structure Training model Label Prediction-> probabilistic values Threshold (0.20)
- 16. Content Define Objective Prepare Data Algorithms Interpret Results Data preparation Feature creation Classification Predictive models 16 Business objective
- 17. Interpret Results • 5-Fold CrossValidation Model performance: + 90% accuracy
- 18. THANK YOU Andrew Ng – Geoffrey Hinton https://www.kdnuggets.com/ https://www.kaggle.com/