Freelance Expert for Predictive Modelling for Boehringer Ingelheim
The global data science team at Boehringer Ingelheim is looking for support by an expert on predictive modelling on a 8-10 week high impact project in early 2016 (Jan – March).
We are looking for an expert on predictive analytics and machine learning. Aim of the project will be to build multiple types of predictive models for the prediction of key financial variablesin the context of clinical development. Data is available and has already been cleaned and consolidated.
Background. First step of the project will be the evaluation of relevant data for predictive modelling by PCA, statistical analysis or other relevant statistical methods. Data enrichment might be required but will most likely be carried out internally.
Secondly, the predictive power of the data for key financial variables will be established and first quick implementations of predictive models will be tested.
Last, multiple predictive modells will be implemented, including random forest models, SVM, deep learning approaches and others. The performance of the models will be evaluated and the best models may then be integrated into a protptype solution.
(There is a potential additional application on time series data, which is currently not in scope however but could lead into a follow-up engagement.)
Collaboration. The freelancer will work closely with the head of the global data science team. The freelancer carry significant responsibility for a high impact project and work very closely with relevant business stakeholders to maximize the outcomes of the project.
The freelancer is expected to pro-actively drive the project and make decisions on analytical and methodological questions with minimal support.
If collaboration is satisfying, follow-up projects are likely; BI Data Science is constantly looking for talented individuals to support the team to manage high project demand. BI Data Science is also currently seeking to grow their team by filling permanent positions with outstanding individuals.
The collaboration will be mostly carried out remotely but will require abour 3-4 visits to Ingelheim am Rhein, Germany (project kick-off, project hand-over, interim workshop). The freelancer will be provided with a BI laptop.
Implementation. The implementation can be done in R or Python (or both). In case the project is succesful, internal dashboard development might be utilized to build a complete first prototype solution.
Data will be provided in form of .csv files. The project is not a big data project.
High performance clusters or environments for parallel computing can be provided if necessary but this is not expected at the moment.
Compensation. Project budget has not been finalized, yet, but compensation will likely lie between 650 – 950 EUR / day.