Profile image of ElanceCoder
Flag of United Kingdom milton keynes, United Kingdom
Member since September, 2012
6 Recommendations

Bosstech Solutions 

Data mining specialist with 4+ year experience of providing data-mining, machine learning, data visualization, data processing, data science, predictive analytic solutions with dedication and thoroughness by meetings the deadlines diligently. Highly skilled in applying various machine learning supervised, unsupervised techniques and algorithms (classification, regression, clustering, association rule mining, decision trees) using multiple predictive analytics software e.g. WEKA, Rapid Miner, Tableau, and R-Programming.
$20 USD/hr
36 reviews
  • 98%Jobs Completed
  • 100%On Budget
  • 100%On Time
  • 15%Repeat Hire Rate

Recent Reviews


Data Analysis using RapidMiner, Weka

Dec 2013

I worked on a confidential longitudinal study of publicly held corporate data for UK 180 companies where I carried out term frequency analyses by text mining and association rules generation in Rapid Miner 5.3. I have wide ranging experience of using SPSS and SAS for data analytics, data visualization and statistical techniques. I also have extensive exposure to Machine Leaning using Rapid Miner, R and Weka that are important online data analysis environments. A significant example of this is my previous experience as working at Shaukat Khanum hospital in Lahore. I was part of a team that applied data mining techniques to patient records in order to provide Doctors with a better understanding of their patient profiles. We took 5 year worth or 60,000 of records with 32 attributes and applied the following models: Data cleaning, Statistical findings, analysis of correlation matrix, visualization by histogram and scatter plotting, frequent item-set mining for visualization of association between attributes. This was important because we can’t find the statistical and visual analytics without applying data mining techniques. We delivered a new perspective to staff at the hospital and as a result Doctors were able to deliver enhanced care to patients delivered in a more precise way that was previously not possible. In my another project I was tasked with analyzing a data set of students to evaluate the relative merits of different techniques for analysis. Students had multiple attributes such as nationality, education, attainment levels and other meta-data. I pre-processed the data then I used Rapid Miner to ascertain confidence via frequent item-set mining to find the associations between all attributes. I analysed Graduate student behaviour using given attributes such as student-major, college, student-level, Max-marks and diploma-description. By using frequent item-set generation I explored the relationship of graduate students to attributes such as residency, citizenship, diploma description and gender. I then created data visualizations, qualitative (histogram-analysis) and quantitative (statistical analysis). I also evaluated classification methods and comparing accuracies by using various sampling techniques and criterias e.g. Gini-index, Information goal and Gain ratio. This was important because it provided me insight of student data set and I found that for some attributes K-Mean cluster performed well and for some attributes decision tree gave highest accuracy. Split validation ‘K Nearest Neighbor’s Shuffled sampling technique gave higher accuracy than other sampling techniques.

Research Assistant

Aug 2010 - Aug 2012 (2 years)

Project worked on:” Analyzing Cancer Hotspots in SKMCH & RC Cancer Registry” Tools Used: Rapid Miner, Microsoft Excel

Junior Web Engineer, assistant manager

Jan 2009 - Sep 2009 (8 months)

Php, My Sql web based coding, assisting project manager in making sales.


MS ComputerScience

2009 - 2011 (2 years)

BS ComputerScience

2004 - 2008 (4 years)


R Programming (2015)


I successfully completed 'R Programming' a course authorized by John Hopkins University and offered through Coursera.