Table 1 Summary of related Papers.

²¹

eSVR, Linear Regression, CNN, STSVR, T-SVR

DEAP Dataset

Proposes a real-time stress recognition framework using peripheral physiological signals.

Blood Volume Pulse (BVP) and Galvanic Skin Response (GSR)

Low prediction error; Suitable for real-world applications

Affected by slight physical movements impacting physiological signals.

²²

Linear Regression, Neural Network

CKD Patient Data

Hybrid model to predict chronic kidney disease using patient data in a cloud setup to enhance smart city healthcare.

Feature Weights (FW)

Improves prediction accuracy over traditional models

Model performance limited by small dataset size.

²³

SVM, K-NN

RALE Lung Sound DB, DEAP Dataset

Compares performance of SVM and K-NN in diagnosing respiratory issues using lung sound signals.

Mel-frequency Cepstral Coefficients (MFCC)

Feature analysis via ANOVA; Comparative classifier insights

Small dataset; Controlled data collection environment.

²⁴

Ranking SVM

NUS-WIDE Dataset

Analyzes user interaction with social images to improve image ranking.

Color, texture, and GIST features

Utilizes robust learning techniques with diverse sensory inputs

Does not consider cultural/geographical image factors.

²⁵

K-NN, AdaBoost, SVM, RF, Logistic Regression

Non-contact Sensor Data

Predicts HR, RR, and HRV from patients in hemodialysis sessions over 23 weeks using non-contact sensors.

Patient age and Body Mass Index (BMI)

High accuracy via machine learning models

Limited in predicting complex clinical events and additional health parameters.

²⁶

Support Vector Machine (SVM)

CRCNSORIG, DIEM

Detects cognitive decline in different age groups using eye-tracking data from video watching.

Pupil size, blink rate, gaze direction, saccade velocity

Enhanced detection with automated feature selection

Few participants; Data gathered in controlled setting.