[Kaggle 보고서] Regreesion with an Abalone Dataset 코드

목차

없음

본문내용

1. Data Loading and Preprocessing
Loaded train.csv and test.csv using the pandas library.
Used LabelEncoder to encode the categorical variable (sex) into numerical values.
Performed feature engineering to create a new feature, size (length x diameter x height).
Cleaned the data by removing rows with a height value of 0..

2. Data Transformation
Applied the np.log1p function to most features to perform log transformation. This transformation helps to make the data distribution closer to a normal distribution, thereby improving model performance.

np.log1p
The np.log1p function from the NumPy library calculates the value of log(1+x). A standard log function log(x) is undefined when x is 0, which can be problematic when taking the log of small values close to 0. The np.log1p function calculates log(1+x), thus allowing for log transformation even when x=0 without any issues.
The np.log1p function is designed to handle values close to 0 safely when performing a logarithmic transformation.

참고자료

· 없음