Primjer 1 - AI Centar Lipik

Proučite dataset gender_classification_v7.csv i definirajte što je target varijabla. Ako je ne pronalazite, izaberite neku varijablu za koju postoje samo dvije vrijednosti u stupcu. Iz dataseta izbacite sve značajke koje nisu numeričke, a ako negdje nedostaje vrijednost umetnite medijan. Model trenirajte sa metodom Naive Bayes. Dataset pripremite sa train_test_split funkcijom i pri tom koristite random_state=42, stratify=y, a 20% neka vam ostane za testiranje. Nakon toga sa preostalih 80% koristite metodu cross_val_score za k-CV pri čemu koristite cv=5 i scoring=‘accuracy’. Na kraju izračunajte točnost modela i sa testnim podacima. Ispišite srednju vrijednost niza scores, ali i grešku sa testnim podacima.

import numpy as np
import pandas as pd
from sklearn.impute import SimpleImputer
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import accuracy_score
from sklearn import model_selection

df = pd.read_csv("gender_classification_v7.csv", low_memory=False)
df.head()

target = df["gender"]
x_num = df.select_dtypes(include=[np.number])

imputer = SimpleImputer(strategy="median")
x_num_imputed = pd.DataFrame(imputer.fit_transform(x_num), columns=x_num.columns)

df_clean = pd.concat([x_num_imputed, target], axis=1)
y = (df_clean["gender"] == "Female").astype(int)
x = df_clean.drop(columns=["gender"]).select_dtypes(include=[np.number])

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=42, stratify=y)

algorithm = KNeighborsClassifier(n_neighbors=3)
ss = model_selection.ShuffleSplit(test_size=0.25)
scores = model_selection.cross_val_score(algorithm, x_train, y_train, cv=ss, scoring="accuracy")

model = algorithm.fit(x_train, y_train)
predictions = algorithm.predict(x_test)
accuracy = accuracy_score(y_test, predictions)

print("Scores", scores)
print("Scores mean", scores.mean())
print("Accuracy", accuracy)

Scores [0.957 0.959 0.965 0.969 0.958 0.97  0.955 0.966 0.96  0.969]
Scores mean 0.9627999999999999
Accuracy 0.973026973026973