Ina Khandelwal
Omega Healthcare Phone: +91-635*******
Bangalore (560008) Email: *********@*****.***
Karnataka, INDIA
Career Objective
To contribute innovatively in a growing organization where there is continuous learning, knowledge sharing culture and scope of development. Seeking challenging opportunities in the field of Machine learning and Deep learning.
Experience
Total experience of around 4 years in Machine Learning and Deep Learning. Machine Learning and Deep Learning Algorithms: Clustering, Classification, Regression, Random walk, ARIMA, Artificial neural networks, Functional Link ANN, Support vector machines, Convolution neural network, Decision trees, Naive bayes classification, Prediction and forecasting models, VGG, LSTM, Natural language processing, Computer vision and Image processing Software Engineer – Data Science Feb 2019
Omega Healthcare, Bangalore
ICD-10 code prediction: Deep learning algorithm for the prediction of ICD-10 codes from patient’s medical report which is the output of OCR (Optical Character Recognition) to automate the process of calculating premium by insurance companies. Developing the model using LSTM for chapter classification and then using CNN with attention mechanism for multiple ICD codes prediction. Senior Machine Learning Engineer Apr 2018 - Dec2018 Fanalytiks, Bangalore
Sentiments analysis for sports data: Predicting sentiments of fans based on twitter posts of users to attract more users and converting them from followers to fanatics. The model is based on LSTM for predicting the sentiments more accurately.
Programming Platform: Jupyter Notebook, Python
Model used: Classification using decision tree, naive bayes and LSTM
Image caption generation model for sports: Developing a image caption generation model for sports related images by extracting features from the image through VGG-19 and generating captions using LSTM model.
Model used: VGG-19 and LSTM
Application Consultant, IBM Nov 2015 – Mar 2018
IBM Certified Cloud Application Developer
Cloud Business Solutions
Time series forecasting using hybrid ARIMA and ANN models based on DWT decomposition: A novel technique of forecasting by segregating a time series dataset into linear and nonlinear components through DWT. The proposed approach tactically utilizes the unique strengths of DWT, ARIMA, and ANN to improve the forecasting accuracy. Programming Platform: Matlab
Model used: Novel model using hybridisation of ARIMA and ANN for linear regression
Forecasting seasonal time series with functional link ANN: Presented the effectiveness of FLANN model for seasonal time series forecasting using unprocessed raw data. Programming Platform: Matlab
Model used: FLANN for regression
Churn prediction using customer engagement score: Predicting churn for website using engagement score as a feature along with page visits, number of sessions, time spent on page, number of hyperlinks visits, time spent on watching video.
Programming Platform: Jupyter Notebook, Python
Model used: Classification using decision tree
Efficient financial time series forecasting model using DWT decomposition: Used DWT to decompose the in-sample training data into linear (detailed) and nonlinear (approximate) components, then applied ARIMA and FLANN model to forecast the respective components. Programming Platform: Matlab
Model used: ARIMA and FLANN for regression
Development of a computer vision assisted system to facilitate head-neck rehabilitation: Developed a computer vision aided system that can be used by patients and physicians to facilitate post injury physical rehabilitation. The system records and analyses the parameters related to movement of head-neck. This helps the patients as well the physicians to assess actual improvement even in the absence of an expert during the rehabilitation process. Language: C++ with OpenCV
A model ranking based selective ensemble approach for time series forecasting: We proposed an ensemble method that selectively combines some of the constituent forecasting models, instead of combining all of them. On each time series, the component models are successively ranked as per their past forecasting accuracies and then we combine the forecasts of a group of high ranked models. Programming Platform: Matlab
Model used: Random walk, ARIMA, ANN, SVM for regression Publications
“Enhancing Time Series Forecasting Accuracy through Multilevel DWT Decomposition” Neurocomputing, (Elsevier) (Communicated).
"Forecasting seasonal time series with Functional Link Artificial Neural Network." Signal Processing and Integrated Networks (SPIN), 2015 2nd International Conference on. IEEE, 2015.
"Time Series Forecasting Using Hybrid ARIMA and ANN Models Based on DWT Decomposition." Procedia Computer Science 48 (2015): 173-179.
"Efficient financial time series forecasting model using DWT decomposition." Electronics, Computing and Communication Technologies (CONECCT), 2015 IEEE International Conference on. IEEE, 2015.
"A Model Ranking Based Selective Ensemble Approach for Time Series Forecasting." Procedia Computer Science 48 (2015): 14-21.
Certifications and Internships
AWS technical essentials training.
IBM certified application developer cloud platform v1.
Internship at IIT Bhubaneswar from 26-05-2014 to 19-07-2014. Academic Records
M.Tech., Computer Science Engineering, The LNM Institute of Information Technology, Jaipur, 2013-2015
(CGPA: 8.56)
B.Tech., Computer Engineering, Govt. Women Engineering College Ajmer, RTU Kota, 2008-2012
(Percentage: 72.54%, Honours)
Senior Secondary Examination, Mangalam D.A.V. Public School Kota, CBSE, 2008 (Percentage: 83.40%) Secondary Examination, Mangalam D.A.V Public School Kota, CBSE, 2006 (Percentage: 87.40%) Technical Knowledge Purview
Programming Languages: Machine learning programming - Python, Matlab, Others - C, C++, Java, Shell
Operating Systems: GNU/Linux, Windows, Mac
Software: Matlab, SPSS, Spark, Jupyter
Others: Latex, SQL, PLSQL, MS-Word/Excel/PowerPoint Libraries: Scipy, Numpy, Pandas, Tensorflow, Keras, OpenCV, nltk, Pytorch Scholastic Achievements
Got offer from UC3M Spain for fully funded PhD and working as researcher.
Recognized for creating an award-winning idea in the IBM’s global Bluemix codefest 2017.
Received IBM manager's choice award thrice.
Successfully completed summer internship project at IIT Bhubaneswar.
Qualified GATE-2013 with 98.02 percentile.
Awarded for securing more than 90% marks in some subjects at school and college.