Untitled

ML

Assignment1: Uber

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
mydata = pd.read_csv('uber.csv')
md['pickup_datetime'] = pd.to_datetime(md['pickup_datetime'])
md.dropna(inplace=True)
corr_mat = md.corr(numeric_only=True)
plt.boxplot(md['fare_amount'])
#removing outliers using interquartile rang
q1 = md['fare_amount'].quantile(0.25)
q3 = md['fare_amount'].quantile(0.75)
IQR = q3-q1
lower_bound = q1 - 1.5 * IQR
upper_bound = q3 + 1.5 * IQR
md = md[(md['fare_amount']>lower_bound) & (md['fare_amount']<upper_bound)]
plt.boxplot(md['fare_amount'])
from sklearn.model_selection import train_test_split
y = md['fare_amount']
x = md[['pickup_longitude','pickup_latitude','dropoff_longitude' ,'dropoff_latitude' ,'passenger_count']]
x_train, x_test, y_train, y_test = train_test_split(x,y,test_size = 0.2)
from sklearn.linear_model import LinearRegression
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import r2_score, mean_squared_error

lr_model = LinearRegression()
lr_model.fit(x_train,y_train)
lr_ypredict = lr_model.predict(x_test)
lr_r2score = r2_score(y_test, lr_ypredict)
lr_rmse = mean_squared_error(y_test, lr_ypredict, squared = False)
print("Linear regression R2 score : ",  lr_r2score)

rf_model = RandomForestRegressor(n_estimators = 100, random_state = 42)
rf_model.fit(x_train,y_train)
rf_ypredict = rf_model.predict(x_test)
rf_r2score = r2_score(y_test, rf_ypredict)
rf_rmse = mean_squared_error(y_test, rf_ypredict, squared = False)

print("Random forest regression R2 score : ",  rf_r2score)
print("Random forest regression RMSE:", rf_rmse)
print("Linear Regression RMSE:", lr_rmse)


Assignment 2: emailspam

import numpy as np
import pandas as pd
data = pd.read_csv("emails.csv")
x = data.iloc[:, 1:3000]
x
y = data.iloc[:,-1].values
y

from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.metrics import accuracy_score

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

# KNN classification
knn_classifier = KNeighborsClassifier(n_neighbors=5)
knn_classifier.fit(x_train, y_train)

knn_y_pred = knn_classifier.predict(x_test)
knn_accuracy = accuracy_score(y_test, knn_y_pred)
print("KNN Accuracy:", knn_accuracy)

# SVM classification
svm_classifier = SVC(kernel='rbf')
svm_classifier.fit(x_train, y_train)

svm_y_pred = svm_classifier.predict(x_test)
svm_accuracy = accuracy_score(y_test, svm_y_pred)
print("SVM Accuracy:", svm_accuracy)


Assignment 3 : gradient decent

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

def objective(x):
    return (x+3)**2

def derivative(x):
    return 2*(x+3)

def gradientDesc(alpha,start,iter):
    x_list = list()
    x = start
    x_list.append(x)
    for i in range(iter):
        g = derivative(x)
        x = x-(alpha*g)
        x_list.append(x)
    return (x_list)

alpha = 0.01
start = 5
iter = 200

x_cor = np.linspace(-15,15,100)
plt.plot(x_cor, objective(x_cor))
plt.plot(start, objective(start), 'ro')

x_list = gradientDesc(alpha, start, iter)
x_cor = np.linspace(-5,5,100)
plt.plot(x_cor, objective(x_cor))
x_list = np.array(x_list)
plt.plot(x_list, objective(x_list),'.-', color ='pink')
plt.show()
print(x_list[-1])


ass 4: k-means clustering

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler

data = pd.read_csv('sample.csv', encoding='unicode-escape')

data.info()

features = data[['QUANTITYORDERED', 'PRICEEACH', 'SALES', 'MSRP', 'YEAR_ID',
'QTR_ID', 'MONTH_ID']]

# Handling missing values (you might want to customise this based on your data)
features.isna().sum()

# scaler = StandardScaler()
# scaled_features = scaler.fit_transform(features)
sse = [] #SUM OF SQUARED ERRORS
for k in range(1, 11):
    kmeans = KMeans(n_clusters=k, random_state=42)
    kmeans.fit(features)
    sse.append(kmeans.inertia_) # model ka error


# plt.figure(figsize=(8, 6))
plt.plot(range(1, 11), sse, marker='x')
plt.title('Elbow Method fomber of Clusters (k)')
plt.ylabel('Sum of Sqr Optimal k')
plt.xlabel('Nuuared Distances')
plt.text(2.5, 0.4, 'Optimal k')
plt.show()

optimal_k = 6 # replace with your chosen k
kmeans_optimal = KMeans(n_clusters=optimal_k, random_state=42)
clusters = kmeans_optimal.fit_predict(features)
print(clusters)

plt.scatter(features['QUANTITYORDERED'], features['SALES'], c=clusters)
plt.title('K-Means Clustering')
plt.xlabel('Quantity Ordered')
plt.ylabel('Sales')
plt.show()