Bijan Ranjbar

I'm experienced with designing, implementing and productizing various classification and clustering methods. I have developed deep (Recurrent) Neural Networks for extraction of text semantics, classifying text and forecasting time series sequences, and I know how to optimize these networks in real world settings.
#NN, #NN-Optimization, #RNN, #LSTM, #GRU, #Attention-Mechanism, #Tensorflow-2, #Word2Vec, #GloVe, #Bert, #Roberta #KNN, #LR, #SVM, #XGBoost, #ARIMA, #K-means, #DBSCAN, #T-SNE

I'm experienced with Amazon Web Services (AWS) for development and productization of Machine Learning projects. I have set up and maintained many training, prediction and analytics pipelines on the cloud. I am familiar with the cost-time trade offs and I know how to optimize settings in the benefit of product owners.
#AWS, #EC2, #EMR, #Glue, #Sagemaker, #ECR, #Cloudwatch, #Step-Function

I have years of experience in working with SQL databases, XML files and large sets of Parquet files. I have the working experience as a full time database administrator, and I am very comfortable with large-scale ETL processes based on both TSQL and Spark technologies. I have built and productized large scale data lakes in corporate settings.
#SQL, #TSQL, #Spark, #Pyspark, #Parquet, #JSON, #XML

I am a master in collecting, cleaning, processing, interpreting and visualizing datasets. I know how to collect valuable information from domain experts and present the analysis results in effective way using various visualization techniques and info-graphics. I use state-of-the-art machine learning methods to extract transformational insights from raw data.
#numpy, #pandas, #seaborn, #BeautifulSoup, #PowerBI, #Tableau, #Gephi, #D3js

During 4 years of my PhD research at department of data science, Maastricht University, I worked on an important project called MiSS (Mining Social Structures). This project was focused on implementing entity recognition on mixed set of noisy structured and unstructured data sources. Based on the extracted named entities, pieces of information had to be combined in order to create correct and meaningful networks. The networks had to be visualized in an interactive and interpretable manner.

• See My PhD dissertation

During 1.5 years of postdoc research at TU Delft I developed novel visualization methods to push forward the available tools for research intelligence, and I helped individual researchers, departments and faculties of TU Delft to position their research within the scientific community. The project was called AIDA (Automatic IDentification of Research Trends), in which, I trained the brilliant researchers and staff members of TU Delft to use the AIDA toolbox.

• Download the AIDA Booklet

At CWTS (Center for Science and Technology Studies) I worked as a data scientist. My responsibilities were administration and maintenance of the center's database servers, supporting staff in carrying out their research successfully and development of relational databases from large sets of evolving raw datasets. I was also involved in performing data analytics projects and providing workshops on the use of visualization tools for bibliometric research.

At Exact, as a data scientist, I used cloud technologies and advanced machine learning methods to solve challenging problems. These problems were provided by various stakeholders, and included classification, clustering, forecasting and data analytic tasks. In this job, I used advanced deep learning methods to address NLP and classification problems, I used Spark-based technologies to build and maintain data lakes, and I used cloud technologies to deliver the product to end users.

Service I

Designing Machine Learning Models

Are you dealing with a classification or clustering problem? Do you have a time series that future values need to be predicted? Do you need to identify implicit sentiments in a large text corpus?
I can help you by designing a prototype or final model with acceptable accuracy. I deliver you the model itself or a working endpoint if you wish.

Service II

Delivering Data Analytics Reports

Do you have a large and/or complex dataset that you need to extract insights from? Is there a correlation or causation that should be validated? Are you looking for a specific visualization based on your dataset?
I can help you by performing a sophisticated analysis based on your domain expertise, and delivering a detailed report in your preferred format.

Service III

Performing Data Transformation

Have you recently received a large XML or JSON file that needs to be transformed into a SQL database? Do you have access to an API that you need to query it and turn the results into an Excel sheet? Do you have access to a large data lake from which analytical records should be generated by using cost and time efficient methods?
I can help you by creating a standard pipeline, designing a proper database and transforming your data in the format most convenient for you.

Service IV

Providing Consultancy and Training

Are you writing a research proposal or business plan that includes Machine Learning applications? Do you need training for your staff over the cloud technologies, database systems or machine learning models available in your organization? Do you want to figure out why your machine learning model is not performing as expected?
I can help you by providing consultancy sessions to discuss technical solutions and estimate time and cost requirements. I can train you or your team based on your needs.