Customer Churn Analysis and Future Opportunities

I am a professional analyst with years of experience of training and tuning machine learning model to real dataset. As you just got the dataset (i) do you know whether the data is clean? (ii) does it has any missing value problem? (iii) do you know how to define your outcome variable (i.e., churn) from the dataset or it is already available in the dataset? (iii) do you have any specific set of algorithms that you want to test or you want to give freedom to decide the appropriate analysis? (iv) do you know what features you want to use in the analysis? These are important for you to take the work forward. As I love to use a variety and advanced set of tools I prefer to use R for modelling instead of Python so let me know Python is a requirement. I need to use Hadoop type platform on a daily basis for various project but my personal experience is sometime the client dataset is only is gigabytes. As you mentioned about some map reduce platform could you please let me know (i) what is the size of the data file and (ii) what is the format so that I can give you a better estimate of time and budget.

Hi dear, I have read your project description and I understood all your point. I have done Master’s degree and currently enrolled in PhD program. I am well versed with all parameters of your project. I want to assure you of top quality delivery. Please contact for further discussion, I am positively waiting for your response.

Hi, Below is my experience, and I can deliver you quality work with in specified or agreed time. Myself Ph.D. in advanced analytics having 10+ years of experience in developing and delivering analytical projects using open source python (including in-memory computing), and can deliver your requirement accompanied by a step-by-step word document, what each line or code means. As, freelancer.com bid does not allow for attaching; cannot showcase my expertise in this area. I am available at freelance chat (click on my name and options), to understand the requirements and explain my approach if required in detail. Waiting for your reply (or) initiation for chat discussion. Regards, Dr. Kumar PM.

I have extensive experience in providing statistical learning models for fortune 100 customers. I have worked on different parts customer churn problem space from survival analysis & optimizing customer support function for new customers to identifying drivers of churn & predicting future churn. I'm well versed in ML techniques ranging from parametric models like linear regression to advance algorithms such as random forest & xgboost with implementation in both python & R. Looking forward to hearing from you!

In case you consider me as an option, I would like to have a brief chat regarding the project. I am confident that I can solve your case. Also,I would like to ask you about the dimensionality of your database , I think that deep learning for your case might lead to overfitting. Thanks, Vasileios Tsakalos

I did data mining for DTAC, Thailand's no. 2 telecomms network exactly on churn (and several other areas related to customer retention). I have also done data mining for other industries and the academe. I am able to use many different software packages. More info: o Biomedical scientist / Assistant Professor, M.Sc. Biochemsitry o Educator (Univ. of Maryland) - taught US military personnel o Bioinformatics/ Data mining of large datasets in the biomedical field and in commercial settings o Multi-award-winning photographer o Scientific and medical writing o Research and editing work for technical writeups o SPSS and R - data analysis and visualization

I won't complicate: I am able to do the work and deliver required resuts. I have a reputation of always closing the deal, and my expirience doesn't lack although I'm new in Freelancer concepts. At the end of it all, you always have to start somewhere on the road I work as a business analyst in IT, I gather data, deliver Customer Deliverable Specification, work in an agile environment, having developers do their job constantly monitoring, testing the product, and tutor the customer at the end of the implementation. Looking forward of working with you

I am an expert in data mining and predictive modeling. I have a PhD in physics and extensive experience in machine learning applied to business. I have developed churn models before using R.

Hello I have a lot of experience working with the Big data and statistical analysis. I worked with churn models for mobile phone operators and can do the high level analysis quickly and efficiently Good luck

My name is Jonathan Beazley. I have over 15+ years experience in Marketing Analytics consulting with top consulting firms and the Fortune 100. My specialties are CRM (like your question), branding, segmentation, customer satisfaction, product mix, etc. Additionally I'm SAS certified in propensity/attrition modeling. What I would propose is that we have a call to discuss your business question (free) and possible options to solve it. I think your best choice here is going to be to use logit to provide propensity scores using R. Many in the industry prefer regression based approaches over black box techniques, because: 1. It provides more insight (e.g., what factors impact attrition the most?), 2. The algorithm can easily be programmed into the database to score new records, 3. Simplicity and model performance. Why is a former consultant using freelancer? I'm getting another degree, and seeing if this could be an avenue to make a little extra cash. Look forward to discussing your business questions with you further. Warm Regards, Jon Beazley

I come from signal processing background, so I think I can do a bit better than simple Machine Learning

I am a chartered statistician and hold a PhD in statistics. the area of interest is my specialty and will like to assist you on this project

Hi I can do this for you. depending on how big your data is(MB, GB, TB, PB) I can use different tech to handle it with different budget and time. for example, if your data is just under GB, I can simply use python/pandas/sk-learn, otherwise i need to use big data solution such as spark/tensorflow etc. Please reply for further discussion and milestone etc. I am senior python developer/data analytist with more than 10 years experience. My tech stack: Big Data, Data Analytics, Machine Learning, Deep Learning Python 2.7/3.5 Numpy/scipy/pandas/matplotlib/sk-learn/nltk/opencv/gensim/keras TensorFlow/DataFlow Hadoop/Spark/Kafka ETL(Data extraction, transformation, load) Web crawling/scraping: scrapy/rapidminer/selenium/phantomjs Data Visualization: d3/bokeh/plotly

I have done work and project in same field. I have experience of working in r , python and scikit tool.I can assure you the most reliable and best results.

A proposal has not yet been provided

Hello, I saw that you mentioned Spark, so my question is o you need parallel computing? MLlib is the Spark (I use Python Spark) equivalent of Sklearn. So, if you need parallel computing I will use MLlib, if not I will use sklearn. As you wish! Thanks, Nicolas