El Wisconsin Diagnostic Breast Cancer (WDBC) dataset es un conjunto de datos de características de imagenes de lunares cancerigenos o no.
Este dataset contiene alrededror de 500 muestras de características extraidas de lunares para determinar cancer de piel.
El dataset esta distribuido de la siguiente forma:
- Label (B: 0, M: 1)
- radius (mean of distances from center to points on the perimeter)
- texture (standard deviation of gray-scale values)
- perimeter
- area
- smoothness (local variation in radius lengths)
- compactness (perimeter^2 / area - 1.0)
- concavity (severity of concave portions of the contour)
- concave points (number of concave portions of the contour)
- symmetry
- fractal dimension ("coastline approximation" - 1)
Caracteristicas significativas que representan importancia a la hora de decidir si un lunar es maligno:
coef std err z P>|z| [0.025 0.975]
------------------------------------------------------------------------------
texture 0.3777 0.063 5.977 0.000 0.254 0.502
area 0.0475 0.010 4.685 0.000 0.028 0.067
smoothness 74.4357 30.948 2.405 0.016 13.779 135.092
compactness 2.4326 19.004 0.128 0.898 -34.815 39.681
concavity 7.4069 7.713 0.960 0.337 -7.711 22.524
concave points 70.1621 27.678 2.535 0.011 15.913 124.411
symmetry 15.1245 10.345 1.462 0.144 -5.150 35.399
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent