Aspiring Data Analyst/Scientist/

Engineer

A post-graduate data analyst with an interest in data pre-processing, cleaning, analyzing, and modeling. Relevant experience in Python Programming, SQL, R, and use of Microsoft tools (Excel, Word, Power BI).

Sixteen months of experience as a Python Developer and Data Curator for building HTML webpages of PDF, DOC, and TXT files using regular expressions (REGEX) in Python.

Building an EA simulator


Created a minimum viable product that emulates the EA simulation feature in FIFA(21), and better predicts the outcome of a particular match, by curtailing certain impactful factors such as the ‘home advantage’.


Upon simulating an entire premier league season multiple times, the created simulator yielded an average accuracy of 85.5% in comparison to the real-world standings of the team.

For additional models and related works, click here


Building predictive models


A patient’s length of stay in the hospital after having their surgery is a very dynamic phenomenon that depends on a lot of factors. But gauging inpatient length of stay and assigning it into tangible terms, such as a group of weeks may help the hospital in establishing a general framework for the patient’s length of stay, regardless of external factors.

The ideal model was found to be the calibrated random forest, yielding accuracy and f1 scores of 79.80% and 79.71% respectively for multiclass variable prediction and 92.97% and 92.95% respectively for binary-class variable prediction.

For additional predictive models (waiting time of patient, etc) and related works, click here

Further works

Performed EDA on the COVID cases of Kuwait, UAE, and Oman from its origin until mid-2021. Used analysis to determine the exponential and logistic growths of all waves of each country. Used SIR models to model a prediction on the different waves of each country. Modified model by adding parameters related to behavioral psychology.

For detailed description of the models, click here

Data evaluation and pre-processing of four stocks with their time-series data were collected from the Yahoo! Finance website: Baron Oil Plc (BOIL.L), Premier African Minerals Limited (PREM.L), Iconic Labs Plc (ICON.L), Lloyds Banking Group Plc (LLOY.L). Segmented the data at various tolerance levels using bottom-up piecewise linear segmentation and employed the "next day forecast" prediction.

For detailed description, click here


Get in Touch

Mohammed Farhaan Ahmed

Email: farhaanzaki@gmail.com

Ph no: +447424743260