Nikhil Muralidhar

                    Twitter: @nikhilm_1   LinkedIn: nikhilmuralidhar

                            Email: nmurali1 AT Stevens Dot Edu

[CV]     [Research Statement]     [Teaching Statement]     [Diversity Statement] 

ScAI Lab News

I am an Assistant Professor in the Department of Computer Science at Stevens Institute of Technology.
At Stevens, I founded and lead the Scientific Artificial Intelligence (ScAI) Lab. The primary research focus at ScAI lab entails developing machine learning models that also incorporate scientific knowledge (along with data) governing a process of interest. This area of research is termed Knowledge-Guided Machine Learning (KGML). 

Our group at SCAIL is especially focused on leveraging scientific domain knowledge to improve model generalization, decision interpretability and reduce the negative effects of data paucity and noise. Most recently, we have been focused on leveraging machine learning techniques to address challenges in physics (specifically computational fluid dynamics) to alleviate the cost of expensive simulations using science-guided machine learning models.    

Our other areas of research interest include: 

I completed my Ph.D (Aug. 22') from the Department of Computer Science at Virginia Tech advised by Prof. Naren Ramakrishnan and Prof. Anuj Karpatne. Prior to starting my Ph.D., I was part of the Big Data and Personalization team at The Washington Post and was primarily focused on developing and maintaining news article recommendation systems and online content optimization tools using multi-armed bandits.

Research @ ScAI Lab

Anomaly detection is a ubiquitous and challenging task, relevant across many disciplines. With the vital role communication networks play in our daily lives, the security of these networks is imperative for the smooth functioning of society. To this end, we propose a novel self-supervised deep learning framework CAAD for anomaly detection in wireless communication systems. Specifically, CAAD employs contrastive learning in an adversarial setup to learn effective representations of normal and anomalous behavior in wireless networks. We conduct rigorous performance comparisons of CAAD with several state-of-the-art anomaly detection techniques and verify that CAAD yields a mean performance improvement of 92.84%. Additionally, we also augment CAAD enabling it to systematically incorporate expert feedback through a novel contrastive learning feedback loop to improve the learned representations and thereby reduce prediction uncertainty (CAAD-EF ). We view CAADEF as a novel, holistic, and widely applicable solution to anomaly detection. Our source code and data are available online (https://github.com/rgopikrishna-vt/CAAD)

Generating flow fields (such as pressure and velocity fields) in 3D space is a fundamental task in computational fluid dynamics (CFD), with applications across a vast spectrum of science and engineering problems. An important class of fluid flow problems in CFD is multi-phase flow, where dispersed solid particles are present in the fluid flow. Despite recent developments in deep learning (DL) for CFD applications, current state-of-the-art is still unable to model 3D flow fields, especially in multi-phase flow settings. 

It is with this goal that we introduce PhyFlow, a novel physics-guided deep learning architecture for modeling 3D multi-phase fluid flows, designed to mimic the popular projection method for solving fluid flows in CFD simulations. We demonstrate that PhyFlow, generates high quality flow fields and yields a 49.61% improvement over other state-of-the-art baselines. We also test the quality of PhyFlow based fields by employing them in downstream tasks like particle drag force prediction and demonstrate state-of-the-art results, improving upon the previous best models by 9.89%. Finally, we demonstrate the consistency of PhyFlow predictions with known underlying physics governing equations. Our source code and data are available online at: tinyurl.com/e5mrerb7 .

Timely forecasting of diseases like influenza, helps health organizations and policymakers by allowing them adequate preparation time for decision making. However, due to the non-trivial dynamics of influenza seasonal progression, effective influenza forecasting still remains a challenge despite increasing research interest. The forecasting challenge is further heightened amidst the COVID-19 pandemic, when the influenza-like illness (ILI) counts are affected by various factors such as symptomatic similarities with COVID-19 and shift in healthcare seeking patterns of the general population. Hence, historical influenza forecasting models are unable to adapt to this new trend. Therefore, we present CALI-Net, a neural transfer learning architecture, that uses novel knowledge distillation techniques to 'steer' a historical disease forecasting model to new scenarios where flu and COVID-19 co-exist. CALI-Net allows this automatic adaptation of ILI forecasts by learning to emphasize learning from either COVID-related signals or from historical forecasting models at appropriate times. Through our CALI-Net framework, we are able to effectively exploit representations from historical ILI forecasting models and historical ILI data as well as the limited COVID-related signals.

Physics-based simulations are often used to model and understand complex physical systems in domains like fluid dynamics. Such simulations although used frequently, often suffer from inaccurate or incomplete representations either due to their high computational costs or due to lack of complete physical knowledge of the system. In such situations, it is useful to employ machine learning to fill the gap by learning a model of the complex physical process directly from simulation data. However, as data generation through simulations is costly, we need to develop models being cognizant of data paucity issues. In such scenarios it is helpful if the rich physical knowledge of the application domain is incorporated in the architectural design of machine learning models. We can also use information from physics-based simulations to guide the learning process using aggregate supervision to favorably constrain the learning process. In this paper, we propose PhyNet , a deep learning model using physics-guided structural priors and physics-guided aggregate supervision for modeling the drag forces acting on each particle in a Computational Fluid Dynamics-Discrete Element Method (CFD-DEM). We conduct extensive experiments in the context of drag force prediction and showcase the usefulness of including physics knowledge in our deep learning formulation.

In recent years, the large amount of labeled data available has also helped tend research toward using minimal domain  knowledge, e.g., in deep neural network research. However,  in  many  situations, data is still limited and of poor quality. Can domain knowledge be useful in such settings? Here, we propose domain adapted neural networks (DANN) to explore how domain knowledge can be integrated into model training for deep networks. In particular, we incorporate loss terms for knowledge available as monotonicity constraints and approximation constraints.  We evaluate our model  on  both  synthetic data generated using the popular Bohachevsky function and a real-world dataset for predicting oxygen solubility in water. In both situations, we find that our DANN model outperforms its domain-agnostic counterpart, yielding an  overall mean performance improvement of 19.5%, with a worst and best-case performance improvement of 4% and 42.7% respectively.

Recent hurricane events have caused unprecedented amounts of damage on critical infrastructure systems and have severely threatened our public safety and economic health. The most observable (and severe) impact of these hurricanes is the loss of electric power in many regions, which causes breakdowns in essential public services. Understanding power outages and how they evolve during a hurricane provides insights on how to reduce outages in the future, and how to improve the robustness of the underlying critical infrastructure systems. In this paper, we propose a novel scalable segmentation with explanations framework to help experts understand such datasets. Our method,CnR(Cut-n-Reveal), first finds a segmentation of the outage sequences based on the temporal variations of the power outage failure process so as to capture major pattern changes.This temporal segmentation procedure is capable of accounting for both the spatial and temporal correlations of the underlying power outage process. CnR also provides a novel explanation optimization formulation to find an intuitive explanation of the segmentation, such that the explanation highlights the culprit time-series of the change in each segment. Through extensive experiments, we have demonstrated that our method consistently outperforms competitors in multiple real datasets with ground truth. We further demonstrate results on real county-level power outage data from several recent hurricanes (Matthew, Harvey, Irma) and show that CnR recovers important, non-trivial and actionable patterns for domain experts, while baselines typically do not give meaningful results.

Cyber physical systems (CPSs) are today ubiquitous in urban environments. Such systems now serve as the backbone to numerous critical infrastructure applications, from smart grids to IoT installations. Scalable and seamless operation of such CPSs requires sophisticated tools for monitoring the time series progression of the system, dynamically tracking relationships, and issuing alerts about anomalies to operators. We present an online monitoring system (illiad) that models the state of the CPS as a function of its relationships between constituent components, using a combination of model-based and data-driven strategies. In addition to accurate inference for state estimation and anomaly tracking,illiad also exploits the underlying network structure of the CPS (wired or wireless) for state estimation purposes. We demonstrate the application of illiad to two diverse settings: a wireless sensor motes application and an IEEE 33-bus microgrid.

Publications

Selected Honors & Awards

Contact: nmurali1 at stevens dot edu