Skip to content

Latest commit

 

History

History
70 lines (64 loc) · 4.9 KB

List of Models.md

File metadata and controls

70 lines (64 loc) · 4.9 KB

List of Models

Benchmark Models

  • simplegeometricmean.py (Geometric Mean of [Store, DayOfWeek, Promo] Groups)
    • Private Score: 0.15996, Public Score: 0.15390
    • Features: Store, DayOfWeek, Promo
  • simplemedian.py (Median of [Store, DayOfWeek, Promo] Groups)
    • Private Score: 0.14598, Public Score: 0.14001
    • Features: Store, DayOfWeek, Promo

scikit-learn Models

  • linearregression-independent.py (Independent Regression Models)
    • Private Score: 0.16939, Public Score: 0.14867
    • Features: Promo, SchoolHoliday, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded)
  • linearregression-independent2.py (Independent Regression Models)
    • Private Score: 0.16405, Public Score: 0.14499
    • Features: Promo, DayOfWeek (one-hot encoded)
  • linearregression-independent3.py (Independent Regression Models)
    • Private Score: 0.16411, Public Score: 0.14511
    • Features: Promo, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded)
  • linearregression-independent4.py (Independent Regression Models)
    • Private Score: 0.16209, Public Score: 0.14587
    • Features: Promo, SchoolHoliday, Year, Month, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), AvgCustStore, AvgCustStoreMonth
  • linearregression-independent-log.py (Independent Regression Models)
    • Private Score: 0.15522, Public Score: 0.13742
    • Features: Promo, SchoolHoliday, Year, Month, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), AvgCustStore, AvgCustStoreMonth
  • linearregression-single.py (Single Regression Model)
    • Private Score: 0.42820, Public Score: 0.42460
    • Features: Store, Promo, SchoolHoliday, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), CompetitionDistance, StoreType (one-hot encoded), Assortment (one-hot encoded)
  • linearregression-single2.py (Single Regression Model)
    • Private Score: 0.26838, Public Score: 0.26738
    • Features: Store, Promo, SchoolHoliday, Year, Month, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), CompetitionDistance, StoreType (one-hot encoded), Assortment (one-hot encoded), AvgCustStore, AvgCustStoreMonth, AvgCustStoreYear
  • randomforestregression-independent-log.py (Independent RF Regression Models)
    • Private Score: 0.14536, Public Score: 0.13748
    • Features: Promo, SchoolHoliday, Year, Month, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), AvgCustStore, AvgCustStoreMonth
  • randomforestregression-single2.py (Single RF Regression Model)
    • Private Score: 0.14726, Public Score: 0.14016
    • Features: Store, Promo, SchoolHoliday, Year, Month, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), CompetitionDistance, StoreType (one-hot encoded), Assortment (one-hot encoded), AvgCustStore, AvgCustStoreMonth, AvgCustStoreYear
  • ridgeregression.py (Independent Ridge Regression Models)
    • Private Score: 0.16179, Public Score: 0.14526
    • Features: Promo, SchoolHoliday, Year, Month, DayOfWeek (one-hot encoded), StateHoliday (one-hot encoded), AvgCustStore, AvgCustStoreMonth

XGBoost Models

  • xgboostregressor.py
    • Private Score: 0.16360, Public Score: 0.14958
    • Features: Promo, SchoolHoliday, DayOfWeek, StateHoliday
  • xgboostregressor2.py
    • Private Score: 0.13205, Public Score: 0.11356
    • Features: Store, DayOfWeek, Year, Month, DayOfMonth, Open, Promo, StateHoliday, SchoolHoliday, StoreType, Assortment, CompetitionDistance, Promo2
  • xgboostregressor-log.py
    • Private Score: 0.12728, Public Score: 0.10754
    • Features: Store, DayOfWeek, Year, Month, DayOfMonth, Open, Promo, StateHoliday, SchoolHoliday, StoreType, Assortment, CompetitionDistance, Promo2
  • xgboostregressor-log2.py
    • Private Score: 0.13264, Public Score: 0.11770
    • Features: Store, Year, Month, YearMonth, Open, Promo, SchoolHoliday, CompetitionDistance, Promo2, CompetitionOpenSinceYear, StateHoliday, DayOfWeek, StateHolidayBinary, StoreType, Assortment
  • xgboostregressor-log3.py
    • Private Score: 0.12511, Public Score: 0.11315
    • Features: Store, DayOfWeek, Year, Month, DayOfMonth, Open, Promo, StateHoliday, SchoolHoliday, StoreType, Assortment, CompetitionDistance, Promo2, AvgCustStore, AvgCustStoreMonth
  • xgboostregressor-log4.py
    • Private Score: 0.12708, Public Score: 0.11219
    • Features: Store, DayOfWeek, Year, Month, DayOfMonth, Open, Promo, SchoolHoliday, CompetitionDistance, Promo2, WeekOfYear, CompetitionOpenInterval, PromoOpenInterval, IsPromoMonth, StoreType, Assortment, StateHoliday
  • xgboostregressor-log5.py
    • Private Score: 0.12305, Public Score: 0.11276
    • Features: Store, DayOfMonth, Week, Month, Year, DayOfYear, DayOfWeek, Open, Promo, SchoolHoliday, StateHoliday, StoreType, Assortment, CompetitionDistance, CompetitionOpenYearMonthInteger, AvgSales, AvgCustomers, AvgSalesPerCustomer
  • xgboostensemble.py
    • Private Score: 0.11880, Public Score: 0.10640
    • Models: Predictions from xgboostregressor-log5 & xgboostregressor-log