Sazee S.'s repositories
Artificial-Intelligence-In-Bioinfromatics-Project1
Computer Science and Information Technology Professionals are being employed in many diverse areas of Science. In this project are focusing on Bioinformatics and the knowledge and skills to understand and participate in this field of research or as an advance step to greater employment opportunities. This project will clarify the fundamentals of Bioinformatics from the end user perspective and will allow us to participate in information gathering.
Big-Data-Manipulation-On-Cloud
I used big data tools (Hive, SparkRDDs, and Spark SQL). I solved challenging big data processing tasks by finding highly efficient solutions. Experienced processing four different types of real data: Standard multi-attribute data (video game sales data), Time series data (Twitter feed), Bag of words data, A News aggregation corpus.
Convolutional-Neural-Networks_MNIST_dataset_image_classification
The project is to predict the value of a handwritten digit, using the very popular MNIST dataset - a collection of 70,000 grayscale handwritten numbers between 0 and 9.
Customer-Churn-Analysis
Customer retention is a critical stage for customer relationship management (CRM), especially for established businesses after their initial exponential growth. Churn management or attrition management is important as when customers leave, there arenegative impacts on revenues. Churn analytics has been widely applied to proactive customer retention where descriptive and predictive analytics are utilised to identify and predict customer propensity to churn.
Customer-Segmentation-and-Profiling
Customer segmentation is a pivotal task for business analytics. Customer segmentation is the process of splitting customers into different groups with similar characteristics for potential business value proposition. Many companies find that segmenting their customers enable them to communicate, engage with their customers more effectively. Future Bank is conducting an analysis on the existing customer profiles and the marketing campaign data to identify the target customers who are mostly likely to subscribe long-term deposits. As a member of the data analytics team, I am tasked to analyse historical data and develop predictive models for marketing purposes. I have used SAS Enterprise Miner and Rstudio to perform the analysis.
GFPGAN
GFPGAN aims at developing Practical Algorithms for Real-world Face Restoration.
handson-ml3
A series of Jupyter notebooks that walk you through the fundamentals of Machine Learning and Deep Learning in Python using Scikit-Learn, Keras and TensorFlow 2.
Image-augmentation-techniques
Data augmentation is the name for the collection of techniques used to increase the amount of usable data. In computer vision this usually means applying various spatial and colour transformations to images. In this project I explored some image augmentation techniques and how they can boost training performance.
Machine-learning-models
Built neural networks (NNs) and Regression models for supervised learning. The NN task is formulated as multi-class classification problem for hand-written images, and the goal is to model the relationship between an image’s content and label. Also uses knowledge on Regression models to predict housing prices in Boston to develop Machine Learning skills.
Predicting-forecast-cover-in-region-using-CNN--Debugging-Neural-Networks
In this project I used cartographic variables of small regions of forest (30m x 30m) to predict the type of forecast cover in each region. Example cover types include: spruce/fir (type 1), lodgepole pine (type 2), etc.
Predictive-Analytics-on-used-cars-data-set
A used car online selling company in the USA is in the process of updating their car price assessment method where they want to apply a data driven technique. The trial dataset consists of 25 variables describing 23531 car sales from 2019 to 2020. The management is very keen to apply predictive modelling for this task where the trail data set is to be used to build and evaluate predictive models to ascertain the feasibility of such an approach. The company outsourced the task to me.
Probability-statistics-using-r
These are projects I did in university during probability and statistics on R studio. I scored 91% in the subject. The projects involve finding probabilities of events, support and confidence of bayes rules. Performed statistical analysis on Groceries data set to find association rules. Also used probability density function, integrate function, joint probability mass function, poisson distribution, probability mass function, binomial distribution, M|M|1 queue and hypothesis testing in R studio. I also wrote an essay summarizing an article written by a statistician (Jim Ridgway) for someone who is not familiar with statistical terms.
Python-programming-language-and-the-Pytorch-machine-learning-framework
In this project, we'll get familiar with the Python programming language and the Pytorch machine learning framework. The combination of Python and Pytorch facilitate rapid machine learning development and experimentation, while also being suitable for production-ready systems. we will: understand Python syntax and control structures be familiar with Pytorch tensor operations
Sentiment-Analysis
As data scientist for the multinational technology company Apple Inc, I developed a sentiment analytics engine for Twitter, which is used to predict consumers’ review sentiments. The aim is to develop both dictionary based and machine learning-based sentiment analytics scripts using a number of R libraries and SAS Sentiment Analysis Studio. I used the developed engine to predict Apple reviewers’ sentiments and benchmark various algorithms and analytics tools.
sql-docs
Technical documentation for Microsoft SQL Server, tools such as SQL Server Management Studio (SSMS) , SQL Server Data Tools (SSDT) etc.
sql-server-samples
Azure Data SQL Samples - Official Microsoft GitHub Repository containing code samples for SQL Server, Azure SQL, Azure Synapse, and Azure SQL Edge
Text-analysis-SAS-Enterprise-Miner-and-Rstudio
I worked as a social marketing analyst in a consulting company to uncover the impacts of online advertising and communication with customers. The aim of the study is to educate the marketing teams of their clients (in diverse industries) to market their products and/or services on social media to maximise customers’ involvement (positive interest and sharing). The company is interested in finding out the relationship between the keywords, comments, sentiments and whether there is a relationship in different topic categories such as entertainment, technology, sports, etc. that are of interest to different clients in various industries.
Text-Sentiment-Analysis-Using-Python
In this project I have separated my code into a number of files, instead of using the notebook to interact with Python- closer to how a real-life project would be structured. To facilitate this via Colaboratory, I have mounted my Google Drive storage to the notebook so I can use it like a regular file system. After this, I moved on to completing a fully functional code-base which allowed me to train models, then save and load them to disk for later usage. I demonstrate this with a text sentiment task - classifying a piece of text as either positive or negative in sentiment.
Transfer-Learning-Image-classification
This project uses transfer learning on a raw dataset collected on a holiday in Africa, where collector went on a safari to observe wild animals in their natural habitat. During this trip he captured 100 photos each of buffalo, elephants, rhinos and zebras. Now they asked me to categorise them. To avoid manually labelling each of the 400 photos, I decide that it would be more fun to build an image classification neural network to solve the problem. This way, I am able to just label 10 images of each animal and let a classifier sort out the rest.
Wine_quality_prediction_using_MLP_Classification_problem
In this project we will convert the problem we solved using regression notebook into a classification problem instead. So we will assign each wine quality rating to a different class. So quality of 0 will belong to a different class to quality of 1. The biggest difference between a regression solution and multi-class classification in terms of implementation is that now our model needs to output 10 values (1 for each quality rating) instead of just a single output. Also we will need to change our loss function.
Wine_quality_prediction_using_MLP_Regression-problem
This project is to predict the quality of wine based on features like alcohol content and density, using a publically available dataset. The target variable of "quality" is a subjective measure of the wine's quality based on expert tasters. I am using pandas, numpy, torch and matplotlib in python.