Uda branvh

project/.ipynb_checkpoints/report-template-checkpoint.md

@@ -0,0 +1,56 @@
+# Report: Predict Bike Sharing Demand with AutoGluon Solution
+#### AJAYI JOHN
+
+## Initial Training
+### What did you realize when you tried to submit your predictions? What changes were needed to the output of the predictor to submit your results?
+TODO: Add your explanation
+Kaggle rejects negative score values, hence, the negative values was set to zero before submission
+
+### What was the top ranked model that performed?
+TODO: Add your explanation
+The top ranked model that performed is the "WeightedEnsemble_L3" with a score  of and RMSE value of -53.055. RMSE evaluation metric was used. Hence, the lower the error, the better the model.
+
+## Exploratory data analysis and feature creation
+### What did the exploratory analysis find and how did you add additional features?
+TODO: Add your explanation
+The temp and atemp features are almost perfectly skewed, indicating that the temperature range is almost the same all through the week. 
+The causal users, windspeed, registered and count columns are left-skewed. This is an indicator that  do not have particular periods of the day to lend a bike.
+The other columns are simply discrete values
+
+### How much better did your model preform after adding additional features and why do you think that is?
+TODO: Add your explanation
+After adding additional features, the model inproved significantly. This included also converting certain features to categorical variables.
+From an initial RMSE value of -53 to RMSE value of -30.1837. Also, the kaggle score improved from 1.79030 to 0.67749
+
+## Hyper parameter tuning
+### How much better did your model preform after trying different hyper parameters?
+TODO: Add your explanation
+Tuning of some of the hyerparameters also improved the model's performace
+Some of the hyperparameters that were tuned included the 'number of boost rounds', 'learning rate',  'num_leaves', 'max_depth'
+The hyperparameter tuning moved the model from a previous RMSE value of -30 to -32.78335 though with a significant increase in the kaggle score wwhich went from 0.67749 to 0.44964. This is probably an indication of the required metric needed for evaluating the model.
+
+### If you were given more time with this dataset, where do you think you would spend more time?
+TODO: Add your explanation
+Given more time with this dataset, I would focus on how working more on the feature engineering and also on the hyperparameter tuning. These two tends to have a significant effect on the model's performance.
+
+### Create a table with the models you ran, the hyperparameters modified, and the kaggle score.
+|model|hpo1|hpo2|hpo3|score|
+|--|--|--|--|--|
+|initial|?|?|?|?|
+|add_features|?|?|?|?|
+|hpo|?|?|?|?|
+
+### Create a line plot showing the top model score for the three (or more) training runs during the project.
+
+TODO: Replace the image below with your own.
+
+![model_train_score.png](img/model_train_score.png)
+
+### Create a line plot showing the top kaggle score for the three (or more) prediction submissions during the project.
+
+TODO: Replace the image below with your own.
+
+![model_test_score.png](img/model_test_score.png)
+
+## Summary
+TODO: Add your explanation

project/report-template.md

@@ -1,39 +1,44 @@
 # Report: Predict Bike Sharing Demand with AutoGluon Solution
-#### NAME HERE
+#### AJAYI JOHN
 
 ## Initial Training
 ### What did you realize when you tried to submit your predictions? What changes were needed to the output of the predictor to submit your results?
-TODO: Add your explanation
+Kaggle rejects negative score values, hence, the negative values was set to zero before submission
 
 ### What was the top ranked model that performed?
-TODO: Add your explanation
+The top ranked model that performed is the "WeightedEnsemble_L3" with a score  of and RMSE value of -53.055. RMSE evaluation metric was used. Hence, the lower the error, the better the model.
 
 ## Exploratory data analysis and feature creation
 ### What did the exploratory analysis find and how did you add additional features?
-TODO: Add your explanation
+The temp and atemp features are almost perfectly skewed, indicating that the temperature range is almost the same all through the week. 
+The causal users, windspeed, registered and count columns are left-skewed. This is an indicator that  do not have particular periods of the day to lend a bike.
+The other columns are simply discrete values
 
 ### How much better did your model preform after adding additional features and why do you think that is?
-TODO: Add your explanation
+After adding additional features, the model inproved significantly. This included also converting certain features to categorical variables.
+From an initial RMSE value of -53 to RMSE value of -30.1837. Also, the kaggle score improved from 1.79030 to 0.67749
 
 ## Hyper parameter tuning
 ### How much better did your model preform after trying different hyper parameters?
-TODO: Add your explanation
+Tuning of some of the hyerparameters also improved the model's performace
+Some of the hyperparameters that were tuned included the 'number of boost rounds', 'learning rate',  'num_leaves', 'max_depth'
+The hyperparameter tuning moved the model from a previous RMSE value of -30 to -32.78335 though with a significant increase in the kaggle score wwhich went from 0.67749 to 0.44964. This is probably an indication of the required metric needed for evaluating the model.
 
 ### If you were given more time with this dataset, where do you think you would spend more time?
-TODO: Add your explanation
+Given more time with this dataset, I would focus on how working more on the feature engineering and also on the hyperparameter tuning. These two tends to have a significant effect on the model's performance.
 
 ### Create a table with the models you ran, the hyperparameters modified, and the kaggle score.
-|model|hpo1|hpo2|hpo3|score|
-|--|--|--|--|--|
-|initial|?|?|?|?|
-|add_features|?|?|?|?|
-|hpo|?|?|?|?|
+|model|CAT|XGB|GBM|RF|score|
+|--|--|--|--|--|--|
+|initial|default|default|default|default|1.79030|
+|add_features|default|default|default|default|0.67749|
+|hpo|'depth':8, 'l2_leaf_reg':10'|'objective':'reg:pseudohubererror', 'eval_metric':'rmse', 'max_depth':10, 'eta':0.03|'objective':'regression_l1', 'num_boost_round':500, 'num_leaves':50, 'eta':0.001, 'random_state'=32|'n_estimators':500, 'max_depth':8, 'min_samples_split':4, 'min_samples_leaf':3, 'max_features':'auto', 'random_state':32|0.44964|
 
 ### Create a line plot showing the top model score for the three (or more) training runs during the project.
 
 TODO: Replace the image below with your own.
 
-![model_train_score.png](img/model_train_score.png)
+![model_train_score.png](img/model_train_score.png)default
 
 ### Create a line plot showing the top kaggle score for the three (or more) prediction submissions during the project.
 
@@ -42,4 +47,4 @@ TODO: Replace the image below with your own.
 ![model_test_score.png](img/model_test_score.png)
 
 ## Summary
-TODO: Add your explanation
+In conclussion, I was able to create a model that can be used to predict the number of persons that makes use of the bike-sharing service in each day of the week. This will help the organization make useful decision and planning, hence, aiding the smooth running of the business.