Projects
Rice Image Detection
In the dataset there are 5 types(or 5 classes) of rice images namely Arborio, Basmati, Ipsala, Jasmine, Karacadag. These contains 15000 images each and a total of 60000 images. The dataset is collected from kaggle. The link of the kaggle dataset iis given below. View Dataset This project is based upon single class image recognition. The model is based upon Convolutional Neural Network. The model predicts that what type of rice is in the given image. The model is build using a tool named tensorflow(by google) and using Jupyter Notebook in Anaconda Navigator.
View Project
Indian Liver Patient Prediction Using Classification
Predicts whether the patient is a liver patient or not. About the dataset. There are some independent variables: (All of these are float type variables) 1. Age 2. Gender 3. Total_Bilirubin 4. Direct_Bilirubin 5. Alkaline_Phosphotase 6. Alamine_Aminotransferase 7. Aspartate_Aminotransferase 8. Total_Protiens 9. Albumin 10. Albumin_and_Globulin_Ratio. Age There is a dependent variable : Dataset (Which is a int type variable consisting of level of the disease). About the model. I used tensorflow for the model creation for Logistic Regression (Best classifier for this dataset). The libraries used are tensorflow, matplotlib, sklearn, numpy, pandas.
View Project
Prediction of Diabetes Using ANN
Predicts whether the patient is a diabetes patient or not.About the dataset. There are some independent variables: (All of these are float type variables). Pregnancies, Glucose, Blood Pressure, Skin Thickness, Insulin, BMI, Diabetes Pedigree Function Age There is a dependent variable : Outcome (Which is a binary type variable) About the model. I used pytorch for the model creation and using a class my_ann. The libraries used are pytorch, seaborn, matplotlib, sklearn, numpy, pandas. I gave 3000 epochs for training of the model. First layer consists of 20 attributes, second layer consists of 20 attributes and the last layer 2 attributes.
View Project