Pravesh Ranchod
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pravesh Ranchod has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
Benjamin Rosman
School of Computer Science and Applied Mathematics, University of the Witwatersrand, Johannesburg, South Africa
2 shared publications- Adedayo Akinade
Carnegie Mellon University Africa, Kigali, Rwanda
1 shared publication - Benjamin Rosman
University of the Witwatersrand, Johannesburg, South Africa
1 shared publication - Birhanu Shimelis Girma
Ethiopian Artificial Intelligence Institute
1 shared publication Bradley Hayes
Department of Computer Science, University of Colorado Boulder, USA
1 shared publication- Branden Ingram
University of Witwatersrand
1 shared publication
Latest Publications
- Open Access
- 7 November 2025
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment
- Vol. 21 (1) , 186-195
Real-time Prediction of Dota 2 Match Outcomes using In-game Chat Logs
This paper investigates the application of supervised learning for the purpose of match outcome prediction from Dota 2 in game chat logs. We analyze a dataset of 50,000 ranked matches, evaluating the predictive power of communication data alone and in combination with game events. Using LSTM and DistilBERT architectures, alongside a logistic regression baseline, we demonstrate that chat logs alone enable accurate prediction (up to 81.4% accuracy), while incorporating game events substantially improves performance (up to 98.4% accuracy). Our temporal analysis reveals that prediction reliability increases significantly during the mid-game phase (15-30 minutes), with models exhibiting different strengths - LSTM achieves higher accuracy while DistilBERT demonstrates greater prediction confidence. This study contributes to esports analytics by establishing chat logs as a viable predictive data source.
- 7 November 2025
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment
- Vol. 21 (1) , 186-195
Open AccessReal-time Prediction of Dota 2 Match Outcomes using In-game Chat Logs
This paper investigates the application of supervised learning for the purpose of match outcome prediction from Dota 2 in game chat logs. We analyze a dataset of 50,000 ranked matches, evaluating the predictive power of communication data alone and in combination with game events. Using LSTM and DistilBERT architectures, alongside a logistic regression baseline, we demonstrate that chat logs alone enable accurate prediction (up to 81.4% accuracy), while incorporating game events substantially improves performance (up to 98.4% accuracy). Our temporal analysis reveals that prediction reliability increases significantly during the mid-game phase (15-30 minutes), with models exhibiting different strengths - LSTM achieves higher accuracy while DistilBERT demonstrates greater prediction confidence. This study contributes to esports analytics by establishing chat logs as a viable predictive data source.
- Open Access
Learning Options from Demonstration using Skill Segmentation
We present a method for learning options from segmented demonstration trajectories. The trajectories are first segmented into skills using nonparametric Bayesian clustering and a reward function for each segment is then learned using inverse reinforcement learning. From this, a set of inferred trajectories for the demonstration are generated. Option initiation sets and termination conditions are learned from these trajectories using the one-class support vector machine clustering algorithm. We demonstrate our method in the four rooms domain, where an agent is able to autonomously discover usable options from human demonstration. Our results show that these inferred options can then be used to improve learning and planning.
Open AccessLearning Options from Demonstration using Skill Segmentation
We present a method for learning options from segmented demonstration trajectories. The trajectories are first segmented into skills using nonparametric Bayesian clustering and a reward function for each segment is then learned using inverse reinforcement learning. From this, a set of inferred trajectories for the demonstration are generated. Option initiation sets and termination conditions are learned from these trajectories using the one-class support vector machine clustering algorithm. We demonstrate our method in the four rooms domain, where an agent is able to autonomously discover usable options from human demonstration. Our results show that these inferred options can then be used to improve learning and planning.
- 1 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 144-149
Raw material selection for object construction
An important step in the construction of novel objects is the ability to recognise combinations of raw materials which are likely to be useful. We aim to exploit the intuition that the visual characteristics of candidate raw materials provide useful cues to their potential combinations. Toward this end, we present a Siamese neural network based model that is able to recognise unseen raw materials present in objects given a list of candidate material images. We demonstrate the utility and efficacy of our model within two domains. The first being a proof-of-concept within Minecraft where we predict the combinations of objects that will result in a target object. The second, more realistic domain, uses the ShapeNet 3D model dataset where we attempt to recover the materials present in a model. We empirically demonstrate that our model is able to learn from a subset of object material pairs and generalise to unseen objects, materials, texture packs. Under these conditions of high visual variation, we show that our model outperforms chance and baseline methods.
- 1 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 144-149
Raw material selection for object construction
An important step in the construction of novel objects is the ability to recognise combinations of raw materials which are likely to be useful. We aim to exploit the intuition that the visual characteristics of candidate raw materials provide useful cues to their potential combinations. Toward this end, we present a Siamese neural network based model that is able to recognise unseen raw materials present in objects given a list of candidate material images. We demonstrate the utility and efficacy of our model within two domains. The first being a proof-of-concept within Minecraft where we predict the combinations of objects that will result in a target object. The second, more realistic domain, uses the ShapeNet 3D model dataset where we attempt to recover the materials present in a model. We empirically demonstrate that our model is able to learn from a subset of object material pairs and generalise to unseen objects, materials, texture packs. Under these conditions of high visual variation, we show that our model outperforms chance and baseline methods.
- 1 March 2016
- Published by IATED Academy
- Vol. 1, 7369-7376
CATCHING PLAGIARISTS: DETECTING PLAGIARISM IN STUDENT SOURCE CODE ASSIGNMENTS IN A VIRTUAL LEARNING ENVIRONMENT
- 1 March 2016
- Published by IATED Academy
- Vol. 1, 7369-7376
CATCHING PLAGIARISTS: DETECTING PLAGIARISM IN STUDENT SOURCE CODE ASSIGNMENTS IN A VIRTUAL LEARNING ENVIRONMENT
- Open Access
Reinforcement Learning with Parameterized Actions
We introduce a model-free algorithm for learning in Markov decision processes with parameterized actions-discrete actions with continuous parameters. At each step the agent must select both which action to use and which parameters to use with that action. We introduce the Q-PAMDP algorithm for learning in these domains, show that it converges to a local optimum, and compare it to direct policy search in the goal-scoring and Platform domains.
Open AccessReinforcement Learning with Parameterized Actions
We introduce a model-free algorithm for learning in Markov decision processes with parameterized actions-discrete actions with continuous parameters. At each step the agent must select both which action to use and which parameters to use with that action. We introduce the Q-PAMDP algorithm for learning in these domains, show that it converges to a local optimum, and compare it to direct policy search in the goal-scoring and Platform domains.
- 1 September 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 471-477
Nonparametric Bayesian reward segmentation for skill discovery using inverse reinforcement learning
We present a method for segmenting a set of unstructured demonstration trajectories to discover reusable skills using inverse reinforcement learning (IRL). Each skill is characterised by a latent reward function which the demonstrator is assumed to be optimizing. The skill boundaries and the number of skills making up each demonstration are unknown. We use a Bayesian nonparametric approach to propose skill segmentations and maximum entropy inverse reinforcement learning to infer reward functions from the segments. This method produces a set of Markov Decision Processes (MDPs) that best describe the input trajectories. We evaluate this approach in a car driving domain and a simulated quadcopter obstacle course, showing that it is able to recover demonstrated skills more effectively than existing methods.
- 1 September 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 471-477
Nonparametric Bayesian reward segmentation for skill discovery using inverse reinforcement learning
We present a method for segmenting a set of unstructured demonstration trajectories to discover reusable skills using inverse reinforcement learning (IRL). Each skill is characterised by a latent reward function which the demonstrator is assumed to be optimizing. The skill boundaries and the number of skills making up each demonstration are unknown. We use a Bayesian nonparametric approach to propose skill segmentations and maximum entropy inverse reinforcement learning to infer reward functions from the segments. This method produces a set of Markov Decision Processes (MDPs) that best describe the input trajectories. We evaluate this approach in a car driving domain and a simulated quadcopter obstacle course, showing that it is able to recover demonstrated skills more effectively than existing methods.