Statistics for Data Analysis

Date: June-July 2013

This course was provided to students in summer-school. After completing this course, each student was required to submit a report on assigned data analytics task.

Contents:

Visualization of data- Plots</strong>: Pie, Bar (horizontal, Vertical, Segmented, Split-chart).
Center Tendency (Mean, Mode, Median).
Variability (Range, Quartiles, Percentile, Variance, Box plot, Standard Deviations).
Correlation and Regression (Correlation, Linear Line fit, best line fit)

FINAL REPORTS SUBMITTED BY STUDENTS - click here

Lecture Handouts: Download

Lec-1: Introduction and Visualization [slides] [handouts]

Lec-2: Measure of Center Tendency [slides] [handouts]

Lec-3: Variability [slides] [handouts]

Lec-4: Correlation and Regression [slides] [handouts]

Lec-5: Further Readings

Introduction to Machine Learning: Supervised and Unsupervised learning, Classification and Regression, Gradient decent, LMS, SVM, Neural Network, Clustering, Recommended System, Anomaly Detection etc,.

Data and Exercises:

Problem: Students marks for Predicting their level of performance Dataset 1 (Download)
For Pivot Table: Download Data file All Results

Links for Online Datasets

http://www.tableausoftware.com/public/community/sample-data-sets
http://calcnet.mth.cmich.edu/org/spss/Prj_carsdata.htm
http://dss.wikidot.com/panel-data-analysis
http://www.kdnuggets.com/datasets/index.html
http://www.models.kvl.dk/datasets (Lot’s of data available on)
http://catalog.data.gov/dataset
http://data.gov.uk/dataset
http://data.gov.in Data about India

Final Topics for Project

After discussion with instructor Fill your final topic for project on following link –>LINK Submit Your Final Topic

Before you choose topic for yourselves Check List of final and approved topics –> List of Approved Topics(New Entry) List of some topics which you can choose –>TOPICS OF CASE STUDIES & PROJECT

Check following submitted Proposal 1. EG14 - Project Proposal Final (Very good proposal by Vivekraj Jayram) 2. EG19 Statistical Analysis of Road Accidents Fatality

Case Studies

Case Study 1 : New Employee's performance Problem Domain A company hires many individuals from various colleges across the country. After a year, they do a comprehensive performance appraisal for all these individuals. The appraisal is based on their on-job performance: a combination of objective metrics met by them and subjective rating by their managers. Each candidate is rated as a best performer (BP), mid performer (MP) or a low performer (LP). During the hiring process, the company had taken these candidates through 15 different tests and an incredible set of profiling. Yet, they landed up with mid and low performers! Full Details of case study - click here to download Download Dataset
Case Study 2 : Classification of Cancer Tissues Problem Domain Cancer is severe disease. Early it is detected more chances to save the patient. There are many expert system that are being developing to classify a patient as Cancer patient or not. Accuracy in this kind of problem is very important. As you don’t want to make even 0.0001% mistake to say a patient that you don’t have cancer. There are many different ways to detect and take observation for same. Finding out the class of tissue can help to make some decision for cancer. Here we have 6 different classes of tissue with 9 observed properties under some experimental situation Full Details of case study - click here to download Download Dataset
Checking Accuracy	Submit your predicted classes of Testing data in following quiz to check accuracy of your system
Case Study 3 :

Final Report submission and Presentation

Your final Report and Presentation should contain

Final Dashboard (as discussed in class, should be connected to all the data)
Report with all the graph and findings (conclusions )
Do not write everything in Presentation Slide which is in report
Email your report to nikesh.14730@lpu.co.in by 26th July ONLY and wait for Review
Take final print out of your file, After reviewed by instructor.
~~One file will be submitted in a group and Presentation will be group wise~~

________________________ FINAL REPORT FEEDBACK!!!

After review of reports, remarks are already given via mail, still some of you have not mailed final report. If you have not received any mail of feedback, do mail again or ask for feedback, probably your mail was not received with final report.

Some of common mistake everybody is doing in report take care of following points while making final report

Stick to format uploaded on UMS, chapter wise and

Do not use too many colors in texts, Colors in Graph and Dashbord is ok, but not in text and in paragraph
Font size and style should be uniform in whole report. (Fixed for Title, Heading and paragraph).(For standard formatting, look any standard book, font size and style and table type is always same in whole book)
Do make your file look like decent, Do not make it too much of tacky
Always use justified paragraph (Ask if not understood)
Provide link for Data and introduction about data in 1st chapter
Use same type of table everywhere for uniformity
Do show your dashbord in report too

Show to any instructor before you take print out

Send me (nikesh.14730@lpu.co.in) Zipped folder of your Report, Dashboard, all Data, All graphs in excel or SAS or SQL and ONE PAGE Summery. Name of folder should be your group number like EG01

You have to give presentation individually so prepare your presentation and report accordingly ________________________

IMPORTANT!!!

Your Final Presentation (Probably) is scheduled individually (Not group wise), So prepare accordingly and bring individual copy of report with yourself. (I tried but it has to be like this)

Still many of you have not sent me ONE PAGE summary of your work. send it ASAP

If you are interested to go for publishing your current work (in journal/Conference) or you want to work on any other topic which can be published, you can join our “DATA ANALYST GROUP” by submitting your details for future contacts on LINK

MANDATORY All of you are Required to submit “ONE PAGE” of your work as Suggested by Loviraj Sir. Prepare that ONE PAGE and mail it to me (nikesh.14730@lpu.co.in) by 26th July. Those who did not attend last class and interaction session with Loviraj Sir are advised to talk to classmates and submit One page of Summary.

Useful links for Data Analytics

Dashboards:

Online: example 1, example 2, example 3, many more, and More

Download : KPI , Dynamic Dashboard, E-commerce, performance, Salary

Reports

Analysis Report by Professionals (UCAS) : Download complete report and go thorough at least once Presentation on Satellite Industry This is Presentation, Not a Report

Links for Writing Report

Link 1, Link 2, Link 3, Link 4, Link 5

Check other Links about Data Analytics and JoB

Salary of Data Analyst in INDIA

References: Chandoo.org, Dundas.com