I’m working as a freelancer of Data analyst and scientist on many projects. Specifically, I’m working in the field of advanced analytics and artificial intelligence while utilizing cutting-edge data science approaches li…
I’m working as a freelancer of Data analyst and scientist on many projects. Specifically, I’m working in the field of advanced analytics and artificial intelligence while utilizing cutting-edge data science approaches like machine learning and deep learning. Experience in data pre-processing, predictive modeling using machine learning and deep learning approaches, statistics, designing data-driven applications and overcoming challenging architecture across several sectors. skilled in data science-related programming languages such as Python, Java, SQL ,…. Delivery involvement spans Healthcare, Financial Services, Public Sector, Natural Language Processing and Computer Vision. I am now looking to excel in a new position within a market leading company as a senior data scientist in an organization, wherein, my technical efficiency and commitment will contribute towards the organizational as well as personal growth.
Data Cleaning and Preprocessing
-Machine Learning and AI Development.
-Data Visualization and Communication.
- Continuous Learning and Research.
Collaboration with stockholder in different domains , including business leaders, domain experts, and IT teams
<p>Analyzed Saudi Telecom customers' sentiments based on tweets to the company's customer care account.<br>
Used the 'customer care tweets KSA' dataset from Kaggle in the Arabic language.<br>
Translated text into English using the googletrans library before annotation.<br>
Annotated sentiments using TextBlob due to the lack of Arabic labeling resources and the absence of sentiment labels in the original dataset.<br>
Cleaned text by removing non-alphabetic characters, links, and usernames.<br>
Split text into tokens.<br>
Lemmatized tokens to get their base forms.<br>
Applied TF-IDF for feature extraction.<br>
Employed Random Forest classifications for sentiment analysis.<br>
Calculated accuracy and confusion matrices to assess classifier performance.</p>
<p>Analyzed public airline sentiment based on the 'Twitter US Airline Sentiment' dataset available in Kaggle.<br>
Performed labeling using TextBlob Data Labeling.<br>
Removed special characters, digits, URLs, and Twitter identities during preprocessing.<br>
Changed all text to lowercase for uniformity.<br>
Tokenized tweets into separate words or tokens.<br>
Lemmatized tokens to return them to their original forms.<br>
Used TF-IDF to vectorize preprocessed text into numerical data.<br>
Used classification models including Nave Base, logistic regression, SVM, neural networks, and Random Forest.<br>
Used accuracy to determine classifier performance in predicting tweet sentiments.</p>
<p>Imported essential Python libraries such as pandas, numpy, and statsmodels.<br>
Plotted the time series to understand trend, seasonality, and patterns.<br>
Used the seasonal_decompose function to break down the time series into trend, seasonality, and residuals.<br>
Used statistical tests such as the Augmented Dickey-Fuller test to check stationarity.<br>
Determined the orders for the SARIMAX model.</p>
<p>Visualized the data with statistical functions.<br>
Created a model for predicting churn to help telecom carriers identify consumers most likely to leave.<br>
Used machine learning techniques including Decision Tree, Random Forest, and SVM.<br>
Obtained the best results by applying Random Forest algorithm.<br>
Demonstrated good accuracy in testing the suggested churn predictive model's classification algorithm.</p>
<p>Collected CSI data for seven distinct human everyday activities.<br>
Transformed CSI data into images.<br>
Used the resulting images as inputs for a 2D Convolutional Neural Network (CNN) classifier.<br>
Demonstrated good accuracy in testing the suggested CSI-based model.</p>
<p>Proposed a two-tier feature selection method to enhance structural-based alert correlation.<br>
Ranked features based on information gain entropy in decreasing order.<br>
Added additional discriminative features beyond the initial ranking.<br>
Evaluated the model using the 2000 DARPA intrusion detection scenario-specific dataset.</p>
<p>Aimed to improve the accuracy of alert correlation.<br>
Tested the proposed model using DARPA 2000 and ISCX2012 datasets.<br>
Evaluated correlation effectiveness using Completeness and Soundness metrics.</p>
<p>Used Convolutional Neural Networks (CNN) to detect objects in the environment.<br>
Used TensorFlow Object Detection API to create, test, and use object detection models.<br>
Included a person and a During in the dataset for the study.<br>
Labeled images manually using Labeling, a graphical image annotation tool.<br>
Exported the inference graph after training.<br>
Used the inference graph containing the object detection classifier.<br>
Transformed CSI data into images and used the resulting images as inputs for a 2D Convolutional Neural Network (CNN) classifier.<br>
Demonstrated good accuracy in testing the suggested model.</p>
<p>Analyzed data and data science in the area of stock market prediction for the next few days.<br>
Conducted interviews with stakeholders to gather input and expectations towards data requirements.<br>
Established the organizational design and timelines to support data analysis operations.<br>
Understood existing processes and business requirements to formulate data modeling recommendations.<br>
Developed a custom tool to help executives and business teams prioritize analytics business use cases.<br>
Visualized the data with statistical functions using matplot library.<br>
Pre-processed data with data cleaning, normalization, handling missing values, and other data transformation techniques.<br>
Combined the Discrete Wavelet Transform (DWT) and Long Short-Term Memory (LSTM) to estimate stock prices in the market for the next few days.</p>
<p>Performed data analysis and data science analysis in the area of CT-image classification of covid-19.<br>
Understood technical requirement and data collection.<br>
Built deep learning model using transfer learning approach (VGG16, VGG19) to predict CT images of Covid 19.<br>
Identified issues, analyzed information, and provided solutions to problems.<br>
Performed image pre-processing including crops, resizing, scaling, and augmentation.</p>
<p>Performed teaching, supervision, and administration work.<br>
Taught Geographical user interface (GIS).<br>
Taught Data Mining and artificial intelligent.<br>
Taught Database Management system.<br>
Taught Programming languages.</p>
The AI-first Recruitment Operating System that unifies and automates the entire hiring lifecycle for modern enterprises worldwide.
Dr Job FZ LLC. 2026 © All Rights Reserved
