Kamini Jodha

**** ***** *****, *** ***,

Fitchburg, WI

ac8vn2@r.postjobfree.com

ac8vn2@r.postjobfree.com

Mob: +1-608-***-****

OBJECTIVE:

To attain a career filled with challenging opportunities & stimulating success in Machine Learning and Data Analysis, where my technical skills & personal abilities can be effectively utilized & improved, eventually leading to contribution in growth of organization. EDUCATION:

Qualification Year Score

Master of Science in Electrical Engineering [University of Wisconsin, Madison]

(Option: Machine Learning and Signal Processing)

2018 3.23

B. Tech Electronics & Communication [Rajasthan Technical University] 2011 78.98% Design Tools

• Programming Languages : MATLAB, Python, Julia, Mathematica, SQL, C

• Libraries Used : Numpy, Pandas, Keras, PyTorch, Tensorflow, scipy

• Cloud platfrom : AWS

• Simulation Tools : Or CAD P Spice, LT spice, Altium Designer Simulation

• PCB CAD : Altium Designer 13

SKILLSET ACQUIRED:

Machine Learning

Regression, Decision Tree, Naïve Bayes Networks, TAN (Tree Augmented Naïve Bayes Networks), Neural Networks, k-Nearest Neighbors, Deep Learning (Convolutional Neural Networks, Recurrent Neural Networks), Support Vector Machines, Natural Language Processing, LSTMs, Distributed Machine Learning for large data. PROJECTS:

UW School of Medicine and Public Health: Applying deep learning techniques and algorithms on medical imaging for generation of synthetic CT from PET/MR scans. (I’m currently working here as a research associate.)

Summer (Industrial) Project on Human Anatomy Detection for PET/MR scans: A directed Summer project for GE Healthcare where a classification of the PET/MR scan slices was to be made using Deep Neural Networks. The classifier can work for both the adults and pediatric dataset. Analysis of various DNNs like ResNets, GoogleNet, VGG, InceptionResnet and Xception was made in terms of accuracy, hyper-parameter tuning, time to train, etc. Also, a brief study of the various activation layers was done using the heat maps. Machine Learning (Academic Projects):

• Stop Sign Detection: A class project for Image Processing training a classifier which detects and gives location for stop sign on a given image.

• Van-gogh Sketches: A class project for Image Processing where multiple ink sketches of Van Gogh’s paintings were provided, which were to be painted using segmentation and classifying parts of an image.

• Attention: A class project where neural machine translation between English and German using the Transformer model described in the paper ‘Attention is All you Need’. The data used is the WMT’14 English German datasets. This Transformer model uses only attention and doesn’t use LSTMs. This makes Transformer model computationally efficient and currently state-of-the-art for Neural Machine Translation tasks.

• Hogwild! Using Distributed Cache: A class project that aims on running Hogwild! on multiple machines with the aid of a distributed cache utilizing a Redis cluster. The motivation behind using a Redis cluster lies in the following gains: it is fault-tolerant, horizontally scalable (which allows addition to the cluster with ease), and it renders an increase in the effective cache size.

• CS760(Machine Learning Course): Implemented learning models like Decision Trees, Neural Networks, Naive Bayes, TAN (Tree Augmented Naive Bayes), kNN without using any library.

• LiDARVoxel: A class project of implementing a neural network to classify objects using raw LiDAR point clouds. Inputs are the Positions of the points in space. Positions and Class labels of objects are the outputs.

Past Work Experience and Relevant Skillset :

Employer : Landis+Gyr Limited, Noida, India

Designation : Hardware Design Engineer

Duration : December 2014 to November 2015

Employer : Secure Meters Limited, Udaipur, India

Designation : Hardware Design Engineer

Duration

: July 2012 November 2014

Hardware Design

Microcontroller & its interfacing with Memory, LCD, Keypad, Transceiver ICs etc.,communication protocols like UART, I2C, SPI, RS485 & Ethernet. Implementation of MODBUS & UART for RS-485 & RS-232 communication. Implementation of I2C & SPI protocol for interfacing EEPROM & FLASH with MCU. Power Supply design for Buck, Flyback topology & Transformer design, PCB design for FR4 materials (2-4 layer boards). BOM, and schematic capture.

Hardware Manufacturing

Design of jigs and fixtures, BONs, test facility development. EMI / EMC tests like RE, CE, ESD, Surge as per IS, IEC & CISPR standards. Exposure to DFM, DFA, FMECA, Agile Framework and CAPA. Hardware:

• Single Phase Electricity Meter – Cost Reduction: Worked on the base design, to reduce cost of the released single phase electricity meter. Also, worked on the RTC distortion customer issue reported.

• Single Phase Electricity Meter & In Home Display: In Home Display, a dedicated display device designed to deliver energy-related information such as energy consumption, pricing, or service messages to / from the energy meter. It is a wired communication device with the electricity meter. It will bring data from the single-phase electricity meter, & is responsible for recharge of meter on wired communication. MSP430F & PIC16F series employed.

• Three phase Panel Meter: A panel meter is an instrument that displays an input signal in either a digital or analog form. Three phase multi-function digital panel meters for accurate and reliable measurement of electrical parameters (voltage, current, power, frequency, etc) in industrial and commercial application. Three phase panel meter features a backlight LCD display that gives excellent read-out of values. Field configurable by push – buttons provided on front of the meter. Designing of modules has been done which is meant for transferring the data being monitored by the panel meter to the external world by external modules such as Ethernet, 4-20mA module. Kamini Jodha Madison, WI

06 – March - 2019

Contact this candidate