Mobile-phones-price-prediction

Used dataset: https://www.kaggle.com/iabhishekofficial/mobile-price-classification

Short description of the dataset:

Feature name	Feature description	Type
id	ID	Numeric
battery_power	Total energy a battery can store in mAh	Numeric
blue	Has bluetooth or not	Boolean
clock_speed	Speed at which microprocessor executes instructions	Numeric
dual_sim	Has dual sim support or not	Boolean
fc	Front Camera mega pixels	Numeric
four_g	Has 4G or not	Boolean
int_memory	Internal Memory in Gigabytes	Numeric
m_dep	Mobile Depth in cm	Numeric
mobile_wt	Weight of mobile phone	Numeric
n_cores	Number of cores of processor	Numeric
pc	Primary Camera mega pixels	Numeric
px_height	Pixel Resolution Height	Numeric
px_width	Pixel Resolution Width	Numeric
ram	Random Access Memory in Megabytes	Numeric
sc_h	Screen Height of mobile in cm	Numeric
sc_w	Screen Width of mobile in cm	Numeric
talk_time	Longest time that battery will last by a call	Numeric
three_g	Has 3G or not	Boolean
touch_screen	Has touch screen or not	Boolean
wifi	Has wifi or not	Numeric

Outcome binary metrics:

Model name	Accuracy real	AUC-ROC real	Average
Logistic regression	98.83 %	98.85 %	98.84 %
Random forest with 80 trees	95.83 %	99.49	97.66 %
Bootstrapped decision tree	95.67 %	99.49 %	97.58 %
Bagging cllassifier	95.00 %	99.43 %	97.22 %
KNN classifier with 98 NN	94.33 %	-	94.33 %
Decision tree	93.50 %	93.53 %	93.52 %
Linear regression	71.69 %	-	71.69 %

Outcome multi-class metrics:

Model name	Accuracy real
Linear regression	91.66 %
Random forest with 250 trees	88.50 %
Decision tree	85.17 %
Logistic regression	79.50 %
KNN classifier with 74 NN	65.00 %
Bootstrapped decision tree	43.67 %
Bagging cllassifier	43.33 %