Data science · Reservoir computing · Nonlinear dynamics
Assistant Professor, Department of Data Science, Amity School of Integrative Sciences & Health, Amity University Haryana
I work at the meeting point of machine learning and dynamical systems: using reservoir computing to learn and predict complex dynamics, and studying how interactions shape collective behaviour in coupled oscillators and self-propelled active drops.
I am an Assistant Professor in the Department of Data Science, Amity School of Integrative Sciences & Health, Amity University Haryana. My research brings together machine learning, particularly reservoir computing, and nonlinear dynamics, with work ranging from synchronization and oscillation suppression in coupled oscillators to experiments on self-propelled active drops. I received my Ph.D. in Physics from the Central University of Rajasthan under Prof. Manish Dev Shrimali and was an Institute Postdoctoral Fellow at IIT Bombay from 2021 to 2023. I have published more than 20 papers in journals including Physical Review E, Chaos, Soft Matter and Europhysics Letters, and I review for Physical Review E, Soft Matter and Chaos.
Learning dynamics with echo state networks, model-free and data-driven prediction using reservoir computing, physics-informed neural networks, and physical reservoir computing.
Synchronization and control, suppression and revival of oscillations, chimera states, critical transitions in coupled systems, and collective dynamics of self-propelled 1-pentanol drops and camphor systems.
Current projects
Machine learning and reservoir computing on landslide records, patient data with blood markers, and EEG recordings.
Predicting the behaviour and collective states of nonlinear systems from data, without a model.
How interactions among groups of three or more units, not just pairs, change synchronization and collective behaviour.
Phase ordering and spatiotemporal patterns in networks of chaotic maps.
Using reservoir computing to predict the nonlinear oscillations of a jumping quarter-car model, with Masahisa Watanabe and Nirmal Punetha.
Research highlights
Machine learning
A parameter-aware reservoir computer built from just two echo-state network units learns the dynamics of a whole star network of oscillators, using the network’s symmetry to keep training cheap.
Active matter
In a closed Y-shaped channel, a self-propelled 1-pentanol drop picks the wider arm 80% of the time and avoids an arm holding a second drop 100% of the time: an artificial form of negative chemotaxis.
Dixit, Chotalia, Shukla, Roy & Parmananda, Soft Matter 19, 6844 (2023)
Active matter
Camphor and pentanol disks moving by Marangoni forces prefer the shortest route to the sink: camphor 80% of the time in a two-path channel and 68% in a three-path one, pentanol 88% and 74%.
Khatun, Chotalia, Das, Dixit & Parmananda, PCCP 26, 25590 (2024)
Coupled oscillators
When oscillators interact only while their mean state lies in part of phase space, oscillations die abruptly, and the oscillating and dead states coexist over a range of coupling near the transition.
Experiments
Experiments on coupled Chua’s circuits show that making the interaction intermittent, in time or by state, can suppress oscillations with attractive coupling or stimulate activity with repulsive coupling, giving access to everything from fixed points to chaos.
Dixit, Aravind & Parmananda, Physical Review E 106, 014203 (2022)
Coupled oscillators
Dynamic coupling in a network of identical oscillators produces synchronization, amplitude and oscillation death, mixed states and bistability; with spatial interaction, initial conditions decide whether the transition is first or second order.
Dixit, Chowdhury, Prasad, Ghosh & Shrimali, Chaos 31, 011105 (2021)
Coupled oscillators
Oscillators that couple to their environment only when their amplitude crosses a threshold, as in biological quorum sensing, show an abrupt first-order transition to a steady state.
Asir, Dixit & Shrimali, Eur. Phys. J. Special Topics 230, 3211 (2021)
Control
Combining drive–response coupling with linear augmentation lets a response oscillator lock to the driver’s frequency while its amplitude is regulated separately, shown for Stuart–Landau and Van der Pol oscillators.
Beyond physics
Grafting and pollination links between fruit trees, modelled as a coupled map lattice, lead to anti-synchronization: half the trees bear heavily each year, so total production stays constant despite alternate bearing.
Chauhan, Dixit, Shrimali, Sakai & Prasad, Chaos, Solitons & Fractals 181, 114639 (2024)
Beyond physics
Modelling each partner’s feeling as a complex number and adding memory of the other’s past behaviour gives multistable, aperiodic relationships whose course depends on how they start.
Self-propelled particles move by creating a surface-tension gradient around themselves (the Marangoni effect). Placed at a junction, a 1-pentanol drop chooses a path: it senses geometric asymmetry in the channel about 80% of the time and chemical asymmetry every time. The blue line is the drop’s tracked trajectory.
Google Scholar, as of October 2026. See current numbers
2026
Prediction of nonlinear oscillations in a jumping quarter-car model using reservoir computing
An echo state network trained on data from only a few points reproduces the period-doubling route to chaos of a quarter-car model whose tyre can lose contact with the road.
Masahisa Watanabe, Shiva Dixit, Nirmal Punetha, Swati Chauhan, Manish Dev Shrimali
arXiv:2609.22205 (2026)
Predicting phase ordering in chaotic maps and coupled map lattices
Predicts how the states of chaotic maps and coupled map lattices order themselves in space, using a data-driven approach.
Shiva Dixit, Swati Chauhan, Manish Dev Shrimali
Europhysics Letters (2026)
Targeting amplitude-frequency of drive-response system through linear augmentation
A hybrid scheme of drive–response coupling plus linear augmentation controls a nonlinear oscillator’s frequency and amplitude at the same time.
Shivani Aggarwal, Shiva Dixit, Awadhesh Prasad, Manish Dev Shrimali
Europhysics Letters 155, 33001 (2026)
2025
Predicting collective states of a star network using reservoir computing
A parameter-aware reservoir computer with only two echo-state network units predicts the dynamical states of a star network of oscillators.
Swati Chauhan, Swarnendu Mandal, Shiva Dixit, Manish Dev Shrimali
Chaos: An Interdisciplinary Journal of Nonlinear Science 35 (12), 123112 (2025)
2024
Constant production in an orchard: an interaction-based approach
Grafting and pollination links in a coupled map lattice of fruit trees produce anti-synchronization, so total orchard production stays constant despite alternate bearing.
Swati Chauhan*, Shiva Dixit*, Manish Dev Shrimali, Kenshi Sakai, Awadhesh Prasad
Chaos, Solitons & Fractals 181, 114639 (2024)
Surface tension gradient invoked path selection
Camphor and pentanol disks moving by Marangoni forces prefer the shortest route to the sink in L-shaped and Y-shaped channels, 68–88% of the time depending on the disk and channel.
Anjuman Ara Khatun, Aarsh Chotalia, Kalpita Das, Shiva Dixit, P. Parmananda
Physical Chemistry Chemical Physics 26, 25590 (2024)
Taming explosive growth through linear augmentation
Linear augmentation turns explosive, unbounded growth in networks of limit-cycle and chaotic oscillators into stable dynamics; the paper also introduces dynamic augmentation.
Shivani Aggarwal, Richita Ghosh, Shiva Dixit, Awadhesh Prasad, Manish Dev Shrimali
Europhysics Letters 148 (2), 23002 (2024)
2023
Pathway selection by an active droplet
A self-propelled 1-pentanol drop in a Y-shaped channel picks the wider arm 80% of the time and avoids a chemically marked arm every time, a form of artificial negative chemotaxis.
Shiva Dixit, Aarsh Chotalia, Shantanu Shukla, Tanushree Roy, P. Parmananda
Soft Matter 19, 6844 (2023)
A memory-based approach to model glorious uncertainties of love
Each partner’s feeling is a complex number; adding memory of the other’s past behaviour makes relationships multistable and aperiodic, so their course depends on how they start.
Aarsh Chotalia*, Shiva Dixit*, P. Parmananda
Chaos: An Interdisciplinary Journal of Nonlinear Science 33, 113131 (2023)
2022
Regulating dynamics through intermittent interactions
Experiments on coupled Chua’s circuits show that intermittent interaction, switched in time or by state, can suppress or stimulate oscillations.
Shiva Dixit*, M. Aravind*, P. Parmananda
Physical Review E 106, 014203 (2022)
2021
Emergent rhythms in coupled nonlinear oscillators due to dynamic interactions
Dynamic coupling among identical oscillators produces synchronization, amplitude and oscillation death, mixed states and bistability, shown for Stuart–Landau, Rössler and ecological models.
Shiva Dixit, S. N. Chowdhury, D. Ghosh, A. Prasad, M. D. Shrimali
Chaos: An Interdisciplinary Journal of Nonlinear Science 31, 011105 (2021)
Dynamic interaction induced explosive death
Oscillators that interact only while their mean state lies in part of phase space show an abrupt, explosive transition to oscillation death, with hysteresis near the transition.
Shiva Dixit, S. N. Chowdhury, D. Ghosh, M. D. Shrimali
Europhysics Letters 133, 40003 (2021)
Critical transition influenced by dynamic quorum sensing in nonlinear oscillators
Oscillators that couple to their environment only above an amplitude threshold, as in quorum sensing, show a first-order transition to a steady state.
P. Asir, Shiva Dixit, M. D. Shrimali
European Physical Journal Special Topics 230, 3211 (2021)
Enhanced synchronization due to intermittent noise
Noise added only in a restricted region of state space keeps chaotic trajectories in the contraction region longer and makes synchronization far more robust than uncontrolled noise.
E. Shajan, M. P. Asir, Shiva Dixit, J. Kurths, M. D. Shrimali
New Journal of Physics 23, 112001 (2021)
2020
Static and dynamic attractive–repulsive interactions in two coupled nonlinear oscillators
Compares fixed and switching attractive–repulsive coupling in a pair of nonlinear oscillators.
Shiva Dixit, M. D. Shrimali
Chaos: An Interdisciplinary Journal of Nonlinear Science 30 (3), 033114 (2020)
Aging in global networks with competing attractive–repulsive interaction
Studies the aging transition, where oscillation is lost as units become inactive, in global networks with competing attractive and repulsive links.
Shiva Dixit, P. Asir, M. D. Shrimali
Chaos: An Interdisciplinary Journal of Nonlinear Science 30 (12), 123112 (2020)
Spatial feedback control on multi-stability in hidden attractors
Applies feedback that acts only in part of the state space to control multistability in systems with hidden attractors.
Shiva Dixit, P. Asir, A. Prasad, N. V. Kuznetsov, M. D. Shrimali
Indian Academy of Sciences Conference Series 3, 13 (2020)
2019
Dynamics of nonlinear oscillator with transient feedback
Feedback switched on only for a time interval or in a region of state space is enough to suppress oscillations in a nonlinear oscillator.
Shiva Dixit, A. Sharma, A. Prasad, M. D. Shrimali
International Journal of Dynamics and Control 7, 1015 (2019)
The dynamics of two coupled Van der Pol oscillators with attractive and repulsive coupling
Maps the dynamics of two Van der Pol oscillators under combined attractive and repulsive coupling.
Shiva Dixit, A. Sharma, M. D. Shrimali
Physics Letters A 383 (32), 125930 (2019)
2016
A fibre-optic sensor for ethanol and methanol based on localized surface plasmon resonance in AgNO₃/ZnO nanorods.
Shiva Dixit, P. Kumar, R. K. Verma
13th International Conference on Fiber Optics and Photonics, OSA Technical Digest, paper Th3A.58 (2016)
2015
Synthesis and structural properties of lead strontium titanate borosilicate glasses with chromium trioxide and graphene nanoplatelets.
C. Gautam, Shiva Dixit, A. Madheshiya
Spectroscopy Letters 48 (4), 280–285 (2015)
Prof. Manish Dev Shrimali, Central University of Rajasthan
15 joint papers, my most frequent collaborator
I am looking for M.Sc. and B.Tech project students, summer and research interns, and Ph.D. students with a background in physics or data science. If you are interested in reservoir computing, machine learning on real-world data, or nonlinear dynamics, email me with your CV and a few lines about what you would like to work on.
| Name | Thesis topic | Joined | Mode | Co-supervisor |
|---|---|---|---|---|
| Neelam | Data-driven mathematical modeling framework for landslide probability prediction | Oct 2024 | Full time | Prof. Paras Chawla |
| Name | Programme | Project | Year | Co-supervisor |
|---|---|---|---|---|
| Anumandla Chandana Priya | M.Sc. Data Science | Modelling and learning biological dynamics: a nonlinear system and reservoir computing approach | 2026 | — |
| Layba Ali | M.Sc. Data Science | Structural feature based connectivity analysis in EEG data | 2026 | Dr. Pitambar Khanra, UPES Dehradun |
| Manish Narwariya | PG Diploma, Data Science | Fake job detection app (summer internship) | 2026 | — |
| Jyoti Yadav | M.Sc. Data Science | Active matter experiments with camphor rotors: image and statistical analysis (research project, manuscript in preparation) | 2025 | Dr. Nirmal Punetha, Amity University Haryana |
| Yujvender Thakran | M.Sc. Data Science | Analysis of deep learning models (LSTM and CNN) | 2025 | — |
| Sahil Kumar Yadav | PG Diploma, Data Science | IPL win probability predictor (summer internship) | 2024 | — |
Amity University Haryana · theory
Amity University Haryana · laboratory
Teaching assistant · IIT Bombay
Teaching assistant · Central University of Rajasthan
Featured talk · July 2024
Gordon Research Conference on Oscillations and Dynamic Instabilities in Chemical Systems
I presented how a self-propelled 1-pentanol drop senses its surroundings: in a Y-shaped channel it picks the wider arm 80% of the time and always avoids an arm holding a second drop. The work was published in Soft Matter (2023), and follow-up results on active disks appeared in PCCP (2024).
Invited talks
Organized
Selected presentations
Awards
Membership
Reviewer for
sdixit@ggn.amity.edu