Discover @zeshan22 Open Source projects

Zeeshan Khalid (@zeshan22)

zeshan22

Stars
5
Global Rank 1,429,452 (Top 50 %)
Followers 1
Registered over 5 years ago
Most used languages

Python
75.0 %

Dart
25.0 %
Location 🇵🇰 Pakistan
Country Total Rank 3,982
Country Ranking

Dart
433

Python
834

chat_app

I had built Fully Functioning Chat App with flutter & Firebase.I had used Firebase Auth, Firebase Firestore, Shared Preference to keep the user logged in, and Stream builder.User can search friends and message them. Chat Room is also created.

Directing-Customers-to-Subscription-Through-App-Behavior-Analysis

The objective of this model is to predict which users will not subscribe to the paid membership, so that greater marketing efforts can go into trying to 'convert' them to paid users. First i applied exploratory data analysis and plotted the histogram of numerical columns , correlation with response variable ,correlation matrix using Matplotlib and Seaborn. I used pandas and numpy for all of our data formatting steps. I build Logistic Regularization model using L1 regularization and find out accuracy of 77%. Our effort has given us model that will label every new user as 'highly likely' to subscribe.

Breast-Cancer-Prediction

Support Vector Machine(SVM) algorithm was able to classify tumors into Malignant/Benign with 97% accuracy. The technique can rapidly evaluate breast masses and classify them in automated fashion. I visualized the data by pair-plot,scatter-plot,count-plot and heatmap with the help of seaborn library. It can be further improved by combing computer vision/ML techniques to directly classify cancer using tissue images.

Student-Performance-Data-Set

The objective of this model is predict student performance in secondary education. First i visualize the grades of each student by all categorical andthe numeric columns. I dropped the columns that does not have great impact on performance. Then i encoding categorical data and applied backward elimination for finding optimal features. I split the data into training and test sets. I trained data on Random forest classifier, Kernel SVM and KNeighborsClassifier. So by confusion matrix and f-score we find out that random forest(77% accuracy) is best classifier for this problem. I plotted feature importance plot we find out absences is the important features for determining the grades of students.