3rd

Summary

4-6 sentences summarizing the most important aspects of your model and analysis, such as:

• The training method(s) you used (Convolutional Neural Network, XGBoost) DecisionTreeRegressor
• The most important features
  ['Open','High','Low','Close']
• The tool(s) you used
  python
• How long it takes to train your model
  127.00032544136047 second

Features Selection / Engineering

• What were the most important features?

We only used four base features and did no other feature engineering

• How did you select features?
  The score on the Public
• Did you make any important feature transformations?
  no
• Did you find any interesting interactions between features?
  No, nofeatureinteraction
• Did you use external data? (if permitted)
  No

Training Method(s)

• What training methods did you use?
  We used a decision tree model, but in the process of parameter tuning, the optimal model was obtained by comparing the scores on the divided validation set

• Did you ensemble the models?
  only one model

• If you did ensemble, how did you weight the different models?
  only one model

Interesting findings

• What was the most important trick you used? We used a decision tree model, but in the process of parameter tuning, the optimal model was obtained by comparing the scores on the divided validation set
• What do you think set you apart from others in the competition? Simple model
• Did you find any interesting relationships in the data that don't fit in the sections above?
  No

Simple Features and Methods

Many customers are happy to trade off model performance for simplicity. With this in mind:

• Is there a subset of features that would get 90-95% of your final performance? Which features? *
  'Open','High','Low','Close' features
• What model that was most important? *
  Tree model
• What would the simplified model score?
  0.352

• Try and restrict your simple model to fewer than 10 features and one training method.
  Only 4 features are used in total

Model Execution Time

Many customers care about how long the winning models take to train and generate predictions:

• How long does it take to train your model?
  111 seconds
• How long does it take to generate predictions using your model?
  277 ms