项目印记

视频网址：https://www.youtube.com/watch?v=i_LwzRVP7bg&t=2835s

数据网址：https://archive.ics.uci.edu/dataset/1/abalone

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

from sklearn.preprocessing import StandardScaler

from imblearn.over_sampling import RandomOverSampler

pd.read_csv("/Users/mutian/downloads/abalone/abalone.data")

cols = ["Sex","Length","Diameter","Height","Whole_weight","Ahucked_weight","Viscera_weight","Shell_weight","Rings"]

df = pd.read_csv("/Users/mutian/downloads/abalone/abalone.data",names=cols)

df.head()

df["Sex"] = (df["Sex"] =="M").astype(int)

df["Sex"].unique

for label in cols[:-1]:

  plt.hist(df[df["Sex"]==1][label],color='blue',label='gamma',alpha=0.7,density=True)

  plt.hist(df[df["Sex"]==0][label],color='red',label='hadron',alpha=0.7,density=True)

  plt.title(label)

  plt.ylabel("Probability")

  plt.xlabel("label")

  plt.legend()

  plt.show()

train,valid,test = np.split(df.sample(frac=1),[int(0.6*len(df)), int(0.8*len(df))])

def scale_dataset(dataframe,oversample=False):

 X = dataframe[dataframe.columns[:-1]].values

 y = dataframe[dataframe.columns[-1]].values

 scaler = StandardScaler()

 X = scaler.fit_transform(X)

 if oversample:

  ros = RandomOverSampler()

  X,y = ros.fit_resample(X,y)

 data = np.hstack((X,np.reshape(y,(-1,1))))

 return data,X,y

print(len(train[train["Sex"]==1]))

print(len(train[train["Sex"]==0]))

Train,X_train,y_train = scale_dataset(train,oversample=True)

train

@@机器学习的基础知识（学习代码记录）