
Aditya Vikram Singh
Data Scientist | Machine Learning Engineer | Data Analyst |
Aditya Vikram Singh is a highly skilled and competent Data Scientist, Machine Learning Engineer, and Data Analyst with over 2 years of experience. He possesses expertise in a range of machine learning techniques including logistic regression, SVM, decision tree, random forest, GBDT, CNN, R
Show more
4.7
Years of Experience
Education
bml munjal university, lucknow public school, saraswati vidhya mandir
Companies
matrixcare, bytelearn, paralleldots, rdiger whrl gmbh, edgistify, applied ai course
Reach out to Aditya Vikram Singh via Email, InMail and SMS drip
by installing Chrome extension
Aditya Vikram's contact details
Email (Verified)
siXXXXXXXXXXXXXXXXXXXXXXom
Mobile Number
+9XXXXXXXXX75
Experience
2022 - Present
matrixcare
Data Scientist
2021 - 2022
bytelearn
Data Analyst
2021 - 2021
paralleldots
Backend Developer
Developed and maintained APIs -> Developed and maintained 20+ APIs for the company's website and business team. The use of APIs by the business team (to get required data) benefited both the tech and business team because the earlier business team used to ask for data from the backend (tech) team. Query Optimization -> Optimized SQL queries for heavy load parts of the system, which improved website responsiveness. Designed Algorithm for Giving Proper Indexing -> Designed and implemented an algorithm that can give proper indexing (shelf number, row number, and stack layer) for products over shelves.
2020 - 2021
rdiger whrl gmbh
Data Scientist
ANPR: Automatic Number Plate Detection (Python, YOLO-V4, CRAFT, CRNN) -> Detecting the license plate number of vehicles of all European countries. Deployed the model on RPi4 and ODroid board. The accuracy of ANPR (deployed on RPi4) is 94% of accuracy. It can detect the all special character of German alphabets and other special characters of European countries. This ANPR is capable of recognizing the country of vehicle by looking at its license plate. Currently working on the premium version of this ANPR. Container Number Detection (Python, YOLO-V4, CRAFT, Classifier) -> Detecting the container number of shipping containers and deployed the model on RPi4 and ODroid board. The overall accuracy of this model is 95%.
2020 - 2020
edgistify
Data Scientist
1. Extracting Warehouse Availability/To-Let Information from Newspapers (Python, YOLO-V4, OpenCV, Pytesseract) -> Extracting the warehouse availability/to-let information from advertisement section of the newspapers. Used custom trained YOLOv4 to locate & find the box around the advertisement sections and then applied images-processing techniques like canny-edge detection, dilation, contours, etc. to find the boxes around each ad. Passed the cropped advertisement section in Pytesseract model for OCR and then applied the text processing techniques to extract the required data. 2. Extracting the Location (i.e. Lat & Long) of Warehouses from Google-Map (Python, YOLO-V4, OpenCV, Selenium) -> Extracting the exact location (i.e. latitude and longitude) of warehouses from Google-Map in a given region of India. Used custom trained YOLOv4 model with manually labeled dataset to locate & find the box around warehouse/s in image of Google-Map. Used Selenium to click at a specific location (obtained from YOLO), to get the location of warehouse from Google-Map and to move the map in a defined region. 3. Building the Warehouse Database (Python, BeautifulSoup, Selenium, Pandas) -> Created the database of the warehouse availability, warehouse details, customer details, agent details for all states of India from various websites.
2019 - 2020
applied ai course
Machine Learning Engineer
1. Toxic Comment Classification and Unintended Bias Reduction (DL/ML, Keras, TensorFlow, Scikit-Learn, Python) -> The dataset has toxic and non-toxic comments. Given a comment model must find the toxicity and reduce the unintended bias. Various ML and DL models (like logistic regression, random forest, Bi-directional GRUs, etc.) were used and then stacked the good performing models. It can be used in online discussion forums. -> Blog: https://medium.com/datadriveninvestor/jigsaw-unintended-bias-in-toxicity-classification-d9adf34307d3 2. Question-Answering System similar to Simple Search Engine (DL/ML, Keras, TensorFlow, Scikit-Learn, Python) -> In the dataset there are 10 answers for each question but only one of them is correct. For a given pair of question and answer model must predict whether it is correct or not. The best model was selected to calculate the MRR (mean reciprocal rank) on test data. Various DL/ML models (like logistic regression, random forest, Bi-directional GRUs, etc.) were used. -> Blog: https://medium.com/@singhadityastudy/simple-search-engine-9062a644da5c?source=friends_link&sk=a2ee7211f6b812a5b8f0039b4c9eb84d
Experience
103 Skills
Amazon Web Services (AWS)
Application Programming Interfaces (API)
architecture
Artificial Intelligence (AI)
Artificial Intelligence (AI)
Backend
BERT
Business Intelligence (BI)
C
Computer Vision
Computer Vision
Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)
Data Analysis
Data Analytics
Data Science
Data Science
Data Scientist
Data Structures
Decision Trees
Deep Learning
Deep Learning
Education
Flask
Flask
GitHub
Gradient Boosting
Heroku
Heroku
Hypertext Transfer Protocol (HTTP)
Image Processing
Image Processing
Java
k-means clustering
Keras
Keras
LightGBM
Logistic Regression
Logistic Regression
Long Short-term Memory (LSTM)
Machine Learning
Machine Learning (ML)
Mathematics
Matlab
matlab
Matplotlib
Microsoft Excel
Microsoft Office
Microsoft PowerPoint
Microsoft Word
MySQL
Natural Language Processing (NLP)
Natural Language Processing (NLP)
Neural Networks
NLTK
NLTK
NumPy
NumPy
Object detection
OpenCV
OpenCV
optimization
Pandas
Pandas (Software)
Probability
Python
Python
Python
PyTorch
Pytorch
R
R
Random Forest
Random Forest
Recommender Systems
Recurrent Neural Networks (RNN)
Recurrent Neural Networks (RNN)
Research Scientist
ResNet
RNN
Scikit
Scikit-Learn
Scikit-Learn
SciPy
scipy
Seaborn
Search
Selenium
Software Engineer
SQL
SQL
Statistics
Supervised Learning
Support Vector Machine (SVM)
Support Vector Machine (SVM)
SVM
TensorFlow
Tensorflow
test
Unsupervised Learning
Vision
XGBoost
XGBoost
Education
2015 - 2019
bml munjal university
Bachelor of Technology - BTech
Mechanical Engineering
2012 - 2014
lucknow public school
Senior Secondary (XII)
2010 - 2012
saraswati vidhya mandir
High School (X)
Colleagues at matrixcare
Show more
Others named Aditya Vikram
Show more
Colleagues at bytelearn
Show more
Colleagues at paralleldots
Show more