Time series forecasting

I have worked with time series as an acoustician and as machine learning scientist.

I have experience with a wide range of time series data: deterministic (acoustic signals), stochastic (prices of financial instruments) and chaotic (non-linear vibration signals).

I have worked with both traditional signal processing methods (including state space methods and feedback and feedforward controllers) and modern Machine Learning methods, such as neural networks (NN), support vector machines (SVM), Gaussian processes (GP), gradient boosted trees (XGBoost, lightGMB and varients).

I have experience using a multitude of neural network architectures for modelling time series, such as wavenets, RNNs with stochastic layers, transformer architectures and I have experimented with using computer vision CNN architectures for working with price trajectories (on price charts).

I have also worked on high dimensional (time series) clustering and similarity search, on covariance matrix estimation and uncertainty quantification. I am very well versed with methods for hyper-parameter tuning and model validation and testing.

A lot of my work in this field is proprietary.

Keywords
Neural Networks, Support Vector Machines, Gaussian Processes, Gradient Boosted Trees, Bayesian Optimization, Wavelets, Embeddings, Entropy, State Space Methods, Nonlinear Control, Chaotic Systems, TensorFlow, Pytorch, Finance