Good And Bad Classification Of Ghugni
Description
Data Description for Good and Bad Classification of Ghugni This project focuses on developing a classification model for the quality assessment of Ghugni, a popular Indian snack made from dried peas or chickpeas, often flavored with spices and served as street food. The dataset contains 1000 samples, evenly divided into 500 "Good" and 500 "Bad" samples. The data is structured to support machine learning classification tasks to differentiate between good-quality and defective batches of Ghugni. Data Overview: Total Samples: 1000 Categories: Good Ghugni: 500 samples Bad Ghugni: 500 samples Features: Sensory Attributes: Appearance: Bright yellow, uniform distribution of spices for good samples; dull color or uneven distribution for bad samples. Texture: Soft, firm texture for good samples; excessively mushy or hard for bad samples. Odor: Pleasant aroma for good samples; rancid or off-smell in bad samples. Chemical Properties: Moisture Content (%): Lower moisture levels for well-cooked samples; higher moisture levels for poorly cooked or spoilt samples. pH Value: Optimal pH indicating freshness for good samples; deviations in spoiled samples. Peroxide Value: Higher peroxide levels indicate rancidity, which is typically found in bad samples. Microbial Analysis: Total Plate Count (TPC) for microbial load Presence of mold/yeast colonies Nutritional Parameters (if available): Protein, fat, and carbohydrate content comparison between good and bad samples. Class Labels: 0: Bad Ghugni 1: Good Ghugni Data Collection Process: Samples were collected from different batches and vendors, ensuring diverse data representation. Quality classification was conducted by trained sensory panels and verified through chemical and microbial analyses. Data Preprocessing: Missing values for chemical data were handled by imputation techniques. Sensory scores were normalized to a common scale. Outlier detection was performed to maintain data integrity. Potential Use Case: This dataset can be leveraged for building supervised learning models such as Logistic Regression, Decision Trees, or Neural Networks for quality control automation in food production. This project aims to ensure consumer safety and consistent product quality by predicting whether a batch of Ghugni meets quality standards.