//assets.cakeresume.com/assets/contentarea/contentarea-2d64eca8dae2a1ffb31e0fb1df887bbad072c7c19734d7d30fa14884bbcc299a.css

An impatient machine learner who is passionate about machine learning and data mining, particularly partial about semi-supervised and reinforcement learning.

University of Essex

Machine Learning Researcher

Ragini Kihlman

ESRC Business and Local Government Data Research Centre, University of Essex 

Infrastructure Developer - Machine Learning

Building a  Federated Learning (FL) prototype system to issues researchers encounter when in the need to use sensitive data stored by companies or public entities and have strict legal and privacy constraints in providing access to the raw data. 
Implemented data-centric as well model-centric federating with vertical learning allowing locally trained models are sent back to the central server after they have been trained on the local devices and are then aggregated (by averaging the weights of all models).

Build a recommender system using collaborative filtering and content-based filtering to optimize what offers to show to each user in their feed and as a sorted list the offers section
Based on the recommendations made and bonuses collected design a learning algorithm that used reinforcement learning to improve the understanding of the company's user base to be able to predict a euro-value on each user, i.e. Customer Lifetime Value (CLV)
Have a paper published depicting the results achieved by the performance of this algorithm
Kihlman, R. and Fasli, M., 2019, December. Augmenting Co-Training With Recommendations to Classify Human Rights Violations. In 2019 IEEE International Conference on Big Data (Big Data) (pp. 2299-2308). IEEE.


Bonusway.fi

Data researcher - Recommendation/Personalisation

Develop robust and efficient analysis framework using Computer vision and matrix factorizations to analyze mouse brain images.
Statistical analysis of mouse brain data using parameterized and non parameterized statistical methods
Analyzing data using Bayesian Analysis with corrected p-value calculations.
Have a paper published explaining the methodology and science involved in this.
Cizeron, M., Qiu, Z., Koniaris, B., Gokhale, R., Komiyama, N.H., Fransén, E. and Grant, S.G., 2020. A brain-wide atlas of synapses across the mouse lifespan. Science.



University of Edinburgh

Informatics Engineer - Data and Machine Learning

Building an infrastructure using machine learning and data mining techniques to streamline the process of victim rehabilitation.
Used active learning with a semi-supervised methodology to identify and categorize survivor stories. 
Have three papers published that talks about the algorithm in details.
Gokhale, R., & Fasli, M. (2017, October). Deploying a co-training algorithm to classify human-rights abuses. In the Frontiers and Advances in Data Science (FADS), 2017 International Conference on (pp. 108-113). IEEE. (Published)
Gokhale, Ragini & Fasli, Maria. (2018). Matrix factorization for co-training algorithm to classify human rights abuses. 2170-2179. 10.1109/BigData.2018.8622397.
Kihlman, R. and Fasli, M., 2019, December. Augmenting Co-Training With Recommendations to Classify Human Rights Violations. In 2019 IEEE International Conference on Big Data (Big Data) (pp. 2299-2308). IEEE.


University of Essex, Freedom from Torture

Senior Research officer

Research in the area of image and signal processing
Successfully converted MATLAB code to C/C++ and CUDA for the MEMD algorithm
Performance testing of the C and C++ version of the Algorithm was 30% faster than the MATLAB code
Performance testing of the CUDA version of the Algorithm was relatively slow due to the version of GPU used.
Identified and suggested new technologies and tools for enhancing the speed and performance of the CUDA version of the Algorithm.


Research Support Assistant 

An excellent 100% track record for on-time and robust implementations, high levels of customer delight, responsible for repeat business from the clients I have worked with, on the backdrop of excellent deliveries
Played a major role in the application maintenance program for a ​ Third party logistics , with effective meeting of the said SLA without breaching the contract.
Led a large multi-location, multi-site implementation programme in the logistics and distribution space which resulted in lowering the day to day tickets and achieving the Application Development contract for the same.
Driven initiatives on reusability & productivity enhancement in the practice
Played the role of an SME and Project manager for the clients’ new initiative for improvement in order fulfilment
processes and implementing the Sterling Multi Channel Fulfilment suite
Involved in functional designs, technical design reviews for multiple complex releases and roll-outs.
Involved in developing team's knowledge and proficiency in Sterling space by conducting, coordinating training sessions and producing and collecting the Knowledge base for Sterling
Client Requirements study/process mapping, driving discussions with Business and IT users
Interaction business on suitable solution to be implemented at the Client’s warehouse locations
Devising a global template-based rollout and support strategy and plans
Providing UI solutions to the enhance the current websites for instant payments, view similar products, promotions
Ensured delivery of a complex solution in constrained timelines with effective coordination with other offshore and
onsite teams.


Share