Department of Applied Mathematics and Theoretical Physics
The Baby-LINC Lab is looking for a Research Assistant/Research Associate interested in Computational Modelling and Machine Learning with strong data manipulation and programming skills.
Candidates should have a Masters or PhD (or equivalent qualification) in Computer Science, Mathematics, Statistics, Neuroscience, Biomedical Engineering or related fields, and be familiar with statistical methods for high-dimensional data, such as latent factor models, regularised regression and classification models, multiple hypothesis testing, and clustering analyses.
The successful candidate will join the fast-paced and dynamic Wellcome Trust LEAP 1kD project on early brain and cognitive development. This interdisciplinary research programme is developing lab-based and portable scalable tools, including dyadic-EEG, eye-tracking, motion-tracking, and speech and video analytics, for the measurement of the effects of social interaction on infants' developing cognition and executive function.
You will join an energetic and friendly international consortium, based at the Department of Applied Mathematics and Theoretical Physics at the University of Cambridge, and work closely with Associate Prof. Victoria Leong at Nanyang Technological University (Singapore) and Prof. Zoe Kourtzi, at the University of Cambridge, (UK).
The primary responsibility of the role holder will be to develop a deep phenotyping computational model of multimodal dyadic sociometric measurements (Leong & Schilbach, 2019; Leong, 2022) collected during parent-child social interactions, in order to predict child executive function (EF). The datasets to be modelled will feature sociometric variables extracted from dyadic electroencephalography (EEG) and electrocardiography (ECG) measurements, dyadic eye-tracking of maternal and infant patterns of eye gaze, dyadic motion and posture of mother and infant, and speech metrics.
The successful candidate may be offered the opportunity to work with data from clinical populations (infants at risk of ASD/ADHD) as well as healthy controls.
About You
Essential:
- Strong knowledge of various machine learning algorithms and statistical methods for high-dimensional data, such as: latent factor models, regularised regression and classification models, multiple hypothesis testing, clustering.
- Proficiency in programming languages like Python or Matlab and in data manipulation and analysis (using relevant libraries, e.g. NumPy, scikit-learn or similar).
Desirable:
- Experience with machine learning frameworks, e.g. JAX, Pytorch, Tensorflow
- Bayesian modelling and computation
- Experience with multi-sensory time series signals and deep learning based approaches
- Experience with modelling temporal data, such as via HMMs, VARs, state space models, GPs
- Experience handling EEG data and data from audio and video sources