Skip to content
Nahid Quader

Papers

Publications

Peer-reviewed papers, newest first. The full list with citations is on Google Scholar.

Master’s research

  1. 2025

    Reinforcement Learning Agent for Multi-Objective Online Process Parameter Optimization of Manufacturing Processes

    Akshay Paranjape, Nahid Quader, Lars Uhlmann, Benjamin Berkels, Dominik Wolfschläger, Robert H. Schmitt, Thomas Bergs

    Applied Sciences (MDPI), 15(13), 7279, 2025

    Manufacturing lines have to balance several conflicting goals at once, such as reducing scrap while keeping the process stable, and the right machine settings have to be chosen in real time while production is running. We frame this as a multi-objective reinforcement learning problem: an agent trained with Multi-Objective Maximum a Posteriori Optimization (MO-MPO) proposes process parameters online. The case study is pinion gear manufacturing, with a high-fidelity Simufact simulation used both to generate training data and to validate the agent. To check that the approach generalises, we also test it on open-source manufacturing datasets and synthetic data. A half-space analysis verifies that the agent’s solutions are Pareto-optimal, and the results show the method is feasible for real-time industrial use.

    Abstract

    Optimizing manufacturing processes to reduce scrap and enhance process stability presents significant challenges, particularly when multiple conflicting objectives must be addressed concurrently. As the number of objectives increases, the complexity of the optimization task escalates. This difficulty is further intensified in online optimization scenarios, where optimal parameter settings must be delivered in real time within active production environments. In this work, we propose a reinforcement learning-based framework for the multi-objective optimization of manufacturing parameters, demonstrated through a case study on pinion gear manufacturing. The framework utilizes the Multi-Objective Maximum a Posteriori Optimization (MO-MPO) algorithm to train a reinforcement learning agent. A high-fidelity simulation of the pinion manufacturing process is constructed in Simufact, serving both data generation and validation purposes. The agent’s performance is assessed using a hold-out test set along with additional simulations of the physical process. To ensure the generalizability of the approach, further validation is performed using open-source manufacturing datasets and synthetically generated data. The results demonstrate the feasibility of the proposed method for real-time industrial deployment. Moreover, Pareto-optimality is verified via half-space analysis, emphasizing the framework’s effectiveness in managing trade-offs among competing objectives.

Undergraduate thesis project

Both papers come from my BSc thesis at BRAC University (2017): a small, basic project that used classical machine learning models to predict whether a movie would be a box-office success, based on data from IMDb, Rotten Tomatoes, Box Office Mojo and Metacritic. It was my first research experience, and my current work is in a different area.

  1. 2017

    A machine learning approach to predict movie box-office success

    Nahid Quader, Md. Osman Gani, Dipankar Chaki, Md. Haider Ali

    20th International Conference of Computer and Information Technology (ICCIT), IEEE, pp. 1–7, 2017

    Abstract

    Predicting society’s reaction to a new product in the sense of popularity and adaption rate has become an emerging field of data analysis. The motion picture industry is a multi-billion-dollar business, and there is a massive amount of data related to movies is available over the internet. This study proposes a decision support system for movie investment sector using machine learning techniques. This research helps investors associated with this business for avoiding investment risks. The system predicts an approximate success rate of a movie based on its profitability by analyzing historical data from different sources like IMDb, Rotten Tomatoes, Box Office Mojo and Metacritic. Using Support Vector Machine (SVM), Neural Network and Natural Language Processing the system predicts a movie box office profit based on some pre-released features and post-released features. This paper shows Neural Network gives an accuracy of 84.1% for pre-released features and 89.27% for all features while SVM has 83.44% and 88.87% accuracy for pre-released features and all features respectively when one away prediction is considered. Moreover, we figure out that budget, IMDb votes and no. of screens are the most important features which play a vital role while predicting a movie’s box-office success.

  2. 2017

    Performance evaluation of seven machine learning classification techniques for movie box office success prediction

    Nahid Quader, Md. Osman Gani, Dipankar Chaki

    3rd International Conference on Electrical Information and Communication Technology (EICT), IEEE, pp. 1–6, 2017

    Abstract

    Movie industry is a multi-billion-dollar industry and now there is a huge amount of data available on the internet related to movie industry. Researchers have developed different machine learning methods which can make good classification models. In this paper, various machine learning classification methods are implemented on our own movie dataset for multi class classification. The main goal of this paper is to conduct performance comparison among various machine learning methods. We choose seven machine learning techniques for this comparison such as Support Vector Machine (SVM), Logistic Regression, Multilayer Perceptron Neural Network, Gaussian Naive Bayes, Random Forest, AdaBoost and Stochastic Gradient Descent (SGD). All of these methods predict an approximate net profit value of a movie by analyzing historical data from different sources like IMDb, Rotten Tomatoes, Box Office Mojo and Meta Critic. For all these seven methods, the system predicts a movie box office profit based on some pre-released features and post-released features. This paper analyzes the performance assessment of all these seven machine learning techniques based on our own dataset which contains 755 movies. Among these seven algorithms, Multilayer perceptron Neural Network gives better result.