BigML Project Ping Yin 2/15/2017
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
BigML ProjectPing Yin
2/15/2017
Reasons for not handling missing value and 0:
• Small portion• Need more background information to impute• 0 means no player in that spot
Split into test and train set
Models
Models
Models
Models
Models
Evaluate models
Evaluate models
Evaluate models
Evaluate models
Evaluate models
Evaluate models
Compare models
Compare models
Compare models
Compare models
Compare models
Compare models
Conclusion: Logistic model is the best model for this dataset
Thanks!
Related Documents
![Florian Tramer` EPFL arXiv:1609.02943v2 [cs.CR] 3 Oct 2016 · PDF filesuch as PredictionIO and cloud-based services offered by Amazon, Google, Microsoft, BigML, and others have arisen](https://static.cupdf.com/doc/110x72/5aa0a0247f8b9a6c178e54f9/florian-tramer-epfl-arxiv160902943v2-cscr-3-oct-2016-as-predictionio-and-cloud-based.jpg)
