Muhammad Asif Khan
I am a data scientist with an excellent knowledge of Machine Learning, Deep Learning Neural Networks, Cloud Computing, Big Data Analytics, Cyber Security, NLP, Advanced Mathematics, and Database Management. I am also a qualified AWS Cloud and ML Practitioner with a solid knowledge of applying security best practices of core AWS services including SageMaker to build, train, tune and deploy ML models.
I have recently completed a Master’s in Computer Science in Big Data Analytics with distinction. During the study, I learned about Linear Algebra, Vectors and Statistics; I am able to understand and apply advanced mathematical machine learning models to solve real life problems.
I am currently seeking to change my professional career from teaching to data science. I am very keen to seek opportunities to solve Big Data problems in finance, cloud computing, cyber security and healthcare.
View My LinkedIn Profile
Portfolio
Research paper on using Artificial Intelligence (AI) in stock market prediction
Evaluate the prediction of Machine Learning models by using real-time financial data
Please press the link below to see the details.
Summary: It was my final dissertation reserach project. The focus of this research was to find out how to use AI and ML to predict the share price of a given stock for an investor to gain the maximum profit. The world financial institutions are generating and accumulating petabytes of data about their customers behaviours; which is commonly known as ‘Big Data’ and it is very important for banks and financial institutions to understand this data, uncover hidden patterns, and produce accurate models to make important decisions accurately. In this study, I employed various Artificial Intelligence (AI) and its subset Machine Learning (ML) and Deep Learning (DL) techniques which have revolutionised and transformed the analysis of Big Data. This study has shown that deep learning technique can be used to understand complex trends and irregular variations in the financial trade data by using the power of its hidden multiple layers and non-linear optimisation kernels.
Developing Artifical Neural Networks (ANN) in PYTHON
Using Deep Learning to understand customers churn rates
Please press the link below to see the details.
Summary: This project is about understanding customers churn rates by using Artifical Neural Network (ANN). I used a sample of 1000 customers and used three layers including input and output layer with a hidden layer. In the first stage, I performed data processing involving few steps like importing the data, encoding categorical data, and splitting the dataset into a training and a test set. The second stage was about creating ANN model consisting of three layers by using approperiate activatation function for each layer. Finally, I compiled the ANN and fit it to the training set by using 100 epochs. In the last satge, I used the test data for the prediction and evaluated the the performance of my model by using confusion matrix.
Applied Machine Learning Project in MATLAB
Using Machine Learning techniques to understand what are the reasons which can affect the heart attack
Please press the link below to see the details.
Summary: I used MATLAB programming environment to build and train various classification models by using the given data. In the first stage, I used pre-processing to fill the missing data and check for outliers. In the second stage, data normalisation was carried out. In the stage three, I used feature selections to focus on those data features that are most likely to produce accurate results by using the Parallel Coordinates Plot. In the stage four, I used three ML classification techniques called Decision Tree, SVM, and KNN to build and train models to solve the given problem. I also evaluated each model’s performance by using percentage accuracy, prediction speed, training time, and AUC. In the last stage, I used performance matrix to comapre three models and used optimisation to select the best model.
Exploratory Data Analysis in Python
To Predict a species of a flower based on its characterstics by using Clustering
Please press the link below to see the details.
Summary: I used Fisher’s Iris data for exploratory data analysis abd used some ML techniques in Python. First, I converted the given data csv file into a dataframe by using pandas. Then, I checked for missing values and dropped all NaN and NA values. I used pandas again to get meaningful statistics and showed statistical significance by using the pairplot with the help of seaborn library. Scatter diagram was also drawn by using matplotlib library. Finally, I used an unsupervised Machine Learning technique called Agglomerative Hierarchical clustering and implement it’s algorithms in Python. I also used the Euclidean distance, and single linkage (minimum distance between any two points) to cluster the given data points by using Dendrogram.
Database Management by using Oracle SQL
Database Designing, Implementating and Testing of a Database
Please press the link below to see the details.
Summary: In this project, I designed a conceptual data model to represent the data requirements of the Help Desk scenario. In my model, I used the Entity Relationship (E-R) diagram to list the attributes of each entity and explain the relationship between various entities. I also discussed the assumptions that I made during the design phase. By using the conceptual model, I developed a relational schema consisting of the definitions of domains and relations that represent the entity and relationship types, including primary keys, foreign keys and any constraints. I used my database design to implement the Help Desk database using Oracle SQL DBMS. I used SQL DDL statements to create tables, insert data and queries to provide the required functionality. Finally, I used SQL DML statements to test the database by answering the selected questions designed by the database administrator.
Game Programming in Python
Tic-Toc-Game
Please press the link below to see the details.
Summary: This is the project I did at the end of the PCEP (Python Certified Entry-Level Programmer certification). This certification is a professional credential that measures your ability to accomplish coding tasks related to the essentials of programming in the Python language. In this project, I used various programming concepts like selection, iteration and using data types. I have demonstrated a solid understanding of resolving typical implementation challenges with the help of the Python Standard Library
Designing a website by using HTML, CSS and JavaScript
Climate Change! Act Now
Please press the link below to view my website.
Summary: This project was to develop a website by using HTML, CSS and JavaScript during the course called ‘Introduction to Web Development’ by Raspberry Foundation. In the first stage, I used HTML to lay out the webpages. Then I used clear CSS to style elements of the webpages consistently. Finally, I used JavaScript commands to make the quiz page more engaging for visitors.
