LAB MANUAL OF DATA VISUALIZATION
AND MACHINE LEARNING
Submitted to
Dr. Raj Kumar
Quantum Ilnivei si , Rooi kee
Towards the partial fulfilment of the degree of
BCA, 2“’ Year
ROORKEE, UNARAKHAND
ACADE3IIC SESSION : 2024 —25
Candidate Faculty In Charge
Name : Nitin Dr. Raj Kumar
Qid : 23120005
Section : B
Index Page
•t•K-means clustering Algorithm
Task: Making images smaller by grouping similar colour
•JLinear Regression
Task: Predicting how much a company will sell waste on, how
much they advertise
4•Decision Tree Algorithm
Task: Scientist classifying Plants and Animals
•«Logisic Regression
Task: Predicting if a customer will cancel a service
'4•Sum :- Support Vector Machine
Task : Recognizing faces (Cats and Dogs)
4 K-NN
Task: Recommending Movies orProduct
â•Random Forest Algorithm
Task: Diagnosis medical condition
•i•Naive Bayes Algorithm
Task: Classify news article
3
K-æœDs clwteFîog Algo7itlua
Task-Making imag sma w @ gFO QÎ@gSÎIOÎÎaF
colour
Python Code
# Generate randbm data
‹›P ndorrœeedt0)
X = npawz¥rn.rarxl(1Œl, 2) û1iXI poäzts ki 2£I
#AppIy KMeans ck›stering k= 3
# numt›er ofcK›sters kmeans =
n_ j
m
0
› Scanned with OKEN Scanner
Linear Regression
task- Predicting how much a company will sell waste
on, how much they advertise
Python Code
# Importing required libraries
import numpy as npimport
matpIotIib.pyplot as plt
from skleam.linearmodel import LinearRagression
# Sample data (adver0sing spend in $1000s and waste sales in tons)
adveñisingspend np.array(t1, 2, 3, 4, s. g. z)).reshape(-1, 1) # X
waste sales np.array([1.5, 2.0, 2.5, 3.5, 3.8, 4.2, 5.0]) #y
# Cr0ate linear regression medalmodel
= LinearRegression()
model.fittadverfising_spend. waste_saIes)
predIcted_salos = model.predict(advertising spend)
# Plotting pILscatter{advertising spend, waste_pales, coIor•’blue‘,
Iabel=’ActuaI Sales')
pft.plc(adveitising_spend. predlcted_saIes, co[or='red',
pit.xIabei('Advertising Spend ($1O00a)') pit.ylabeI(Wa
Sales (Tons)’)
pI1tit/e(’Linear RegfeBBjon: Advertising vs Was, Sales’)
piLlegsnd() piLgrid{True)
› Scanned with OKEN Scanner
Decision Tree Algorithm
Task : Scientist classifying Plants and Animals
Python Code
# lmport necessary libraries from
sklearn import tree import
matpIotIib.pypIot as pit
II Sample dataset: [has leaves, can move, has eyes]
# 0 = No, 1 = Yes
# Features: [has leaves, can move, has eyes\ X
[1, 0, 0], # Plant
[1, D, 0], # Plant
}0, 1, 1], 4 Animal
[0, 1, 1], # Animal
(1, 0, 0], # Plant
[0, 1, 11. # Animal
# Labels: 0 Plant, 1 Animal y
= [0, 0, 1, 1, 0, 1]
# Train Decision Tree Classifier
clf —tree.DecisionTreeCIassifier()clf
clf,fit(X, y)
Plat the tree plt.figure(figsize=(8,6)) tree.plot tree(cIf,
feature_names'["has_Ieaves can_move
"has eyes"],
pit.titIe(' Des s Tree: Plan( VS Animal Cla
plt.show{)
Output
has leavœ <= 0.5
ginì - 0.5
samples = 6
value = [3, 3]
class = Plant
True False
Lo«gistic Regression
P›zftoa Code
lmpon ui lñ ins-
numpya>no pan
modeLprodict_proba(x_t 'snt:. *łt
s
Scanned with OKENScanner
Sum :- Support Vector Machine
Python Co‹lc
# Import necessary libraries
import numpy as nd \mpor\
matplotIib.pyplot as plt from
sklearn import svm
# Sample feature date: ear size, snout length]
# Values are mads-up for demonst‹a\ion
X = np.array([
|2, 1j, # Cat
t1.5, 1}, # Ca\
(2, t.2}, # fat
|8, 6], # Dog
[9, 6.5j, # Oog
l+ 5, 5.8] I' Dog
// Labels: O Cat, 1 = Oog y
{0, 0, 0, 1, 1, 1]
# Train SVM model clf =
svm.SVC{ke ne\=’linear')
#Plotting decision bDundary
pit.figure(figsize=(g,6))
Plot points
pit.scatter|xl:,o),xl:,i),c-y,cmap='coolwarM'.s-i0 %d£ecolnrs-'t'l
it Plot decision boundary
at=ptt.gca{}xlirn=ax.get_zl rnli
Sc
# Create grid to evaluate model
gsnp.Iinspace(xIim(0],xIimtq)
g-pp,|jnspace(yIim[0],yIirn(1])
yy, =np,meshgrid(yy,xx)
gnp,vstack((XX.raveI(),YY.ravE'I())).T
/-df.decision function(Xg).reshape(XX.shape)
r Nat marginsand boundary
tit.‹ontouF(XX,W,Z,COffJ‹s 'k’,fevefs=[-1,0,1},linestyles-('—’,'-’,'1)
p|t.xIabe1('Ear Size') pit.yIabeI('SnoutLengh')
plt.fitIe('SVM: Cat vsDog£ace Classification')
pit.grid(True)
pIt.ShoW()
Output
SVH. Cat vsDogFace CbssifJcaticr
Scanned wi
9 K-NN
Task : Recommending Movies or Product
Pythun Code
# Impart librarles
import numpy as np import matplotIib.pyplot as plt
I\am sklearn.neighbors import KNeighborsClassifier
# Sample daia: [Action rating, Romance rating]
X = np.arrays{
[5, 1]. # Likes action Only
|4, 1]. # Mostly action
(J, 5], # Likes romance only
[2, 4], # Mostly romance
[3 3}, # Balanced
[4, 2], # More action
(1, 4], # More romance
I)
// Labels: 0 = Recommend Action Movie, 1 = Recommend Romance
Msvie
y np,array({0, 0, 1, 1, 1, 0t 1])
# Train K-NN model
knn = KHeighbo‹sCIass ifiertn neighbors=3) km.fit(X,
# Plotting decision bOU*ld I J
h = D.1 x min,
x max o. 6 y min.
I max = 0, 6
Scanned
x yy=np,me5hgéd(np.arang•
np.a f a nQ e (_+"'
. plt.figure(figsize=(8, 6))
lgC0ntAUff(XX, , Z, cmap=ptt.cm.Pastell, alpha=O.6)
} p!t.scatter(X[:, 0], X[:, 1], c—y, cmap=pIt.cm.coolwarm, edgecotors-
s=100)
R”plLxlabel('Action Rating') pIt.ytabe\( Romance
ph.title(‘K-NN: Movie Recommendation System')
Spit.grid(True) plt.show(}
•,.Dutput
Novic Recommendation System
S
Random Forwf Algorithm
joy Code
plmport required libraries
y x›rtnumpy as npimport
[36.6. 70}. // HaaltHy
[37.a, 7Z{, # Henttt‹y
|39.0, 95], # Sick
|36.7, 6B], P Heah y
[38.0, 85], # Sick
(37a 75j, # Healthy
[39.5, 98], # Sick
8Labea: 0=Healthy. 1= Sity
=np y([0. 0, 1, 1, 0, 1. 0, 1]) #
TfginRandomForest m0del
=RandomForestCi» ifier(n estimators-10, random@tate=42)
**btân9 ascision boundary
min, x _ 3g, y min. y_m
yy -Up.rriesñgfid(np.a/ange(x m
*_ h),
"
- Mhspe{mshape)
pit.figure(figsize=(s, 6))
*! TI*P"P!t C/Ti.RdYlBu, alpha=0.6)
pit 1}, c=y, *map=plt.cm.bwr,edgecolors=’k'. s=100)
pit.xlabeI('Temperature(•C)’) It.ylabel(’Heart Rate (bpm)')
bit.title(’RandomForest: I Diagnosis‘
pjt,grId(True} pit.show()
Randomror•st: Medical Ofagnosis
2’ sk : Classify news article
y jmpor! necessary librar es
inportnumpy as np impDrt
a yb.pypJot as plt fram
stl6Brn.naivebayes impurt
Ga ssianN3
r Sarriple fuature data: [sports words count, pofitics words count] X
[5 1]. #ćports
[6 2], # Sportu
{t. 6}, ąPc›/i i«s
|2, 7], g Politics
[4, 1j, # Sports
{I, 5ł. # Poliiics
[ó, 2J, # Sports
l2 6j, // Polłt‹cs
i)
# Labe/s.g- gpgrłs - pqlitics y
* np.array({0, 0, 1, 1. 0, 1. 0, 1])
* Train Naive Bayes modelmodel
' GaussianNB() model.fit(X,y)
P|0hing decision doundary = 0,8
nned with
-y,reshape(xx.shBpl°')
figure(figsize—(8, 6)d
C
*!P'|0lt.CfTl.Pas
tet2, alpha=0,6)
p›cafe(X[:,0],X{:,
Cmap Pvt.¢M.coolwarm,
s-100)plt.xIabel('S 5BOQ 9
Cou nt') p/t,yfabeI('coliti
want) sword
II,title(’Naive Bayes:News Article ClaSS/
fiCdtion')
&.gri4(True) pit.show()
Scanned with OKEN 5
