Salman Khaliq's Linkedin Analytics

👇 My Story 👇 ⚡ My Machine Learning journey 🚗 started in 2008 when I learned to code in MATLAB, and then started implementing the code for the numerical methods course the next year. That's when I joined Dr. Usmani's Lab, where the research focus was on applying machine learning techniques for predictive modeling of time series 🕝 data. ⚡ Later that year our research paper was accepted for oral presentation at the ICESE conference, the topic of the research was "Rainfall 🌧 Forecasting: A comparison between Linear Models and Neural Networks 🧠." We collected the weather data for the past 60 years from the Meteorological Department of Pakistan. The data had 29 features and 1 target variable, we followed the standard steps of EDA and Data Preparation and then ended up using PCA for dimensionality Reduction and then optimized the Multiple Linear Regression model and an Artificial Neural Networks with 1 hidden layer. ⚡ In 2010, I designed my senior year project for predicting Heart Attacks 💔, for which we collected the data of 2006 patients from the Pakistan Atomic Energy Medical Commission. The data contained 60+ features and the target variable had 2 classes and 3 labels within each class. Our preprocessing took the majority of the time for this data. After tidying up, imputing, and encoding the data we ran several classifiers models to predict the classes and labels. This research was also accepted for presentation at ICCAD Conference. ⚡ In 2011, I supervised two senior year projects in my Alma Mater for both projects real-life data from different hospitals were collected and we developed predictive models for ECG results and EMG signals. This year I also worked on a project with several class fellows where we developed a Shortest Path Searching Algorithm for PeopleBot in structured environments. This was published in IEEE Conference ICMLC and Elseviers Journal Procedia Engineering. ⚡ In 2012, I joined LUMS University as a Research Assistant in the Mathematics Department, where I designed mathematical models mimicking the inhibitory behavior of neurons 🧠 using MATLAB. Since neurons communicate electrically and all ion channels can be modeled using partial differential equations. ⚡ In 2018, during my MS I designed a mathematical model mimicking lucid behavior in Alzheimer's patients on MATLAB using the Hodgkin-Huxley model. I presented this model on Cell Biology and Neuroscience Retreat at Rutgers. ⚡ What I am up to right now check out my LinkedIn Profile or just reach out to me.