-
Adaptive Sampling for Probabilistic Forecasting under Distribution Shift
Authors:
Luca Masserano,
Syama Sundar Rangapuram,
Shubham Kapoor,
Rajbir Singh Nirwan,
Youngsuk Park,
Michael Bohlke-Schneider
Abstract:
The world is not static: This causes real-world time series to change over time through external, and potentially disruptive, events such as macroeconomic cycles or the COVID-19 pandemic. We present an adaptive sampling strategy that selects the part of the time series history that is relevant for forecasting. We achieve this by learning a discrete distribution over relevant time steps by Bayesian…
▽ More
The world is not static: This causes real-world time series to change over time through external, and potentially disruptive, events such as macroeconomic cycles or the COVID-19 pandemic. We present an adaptive sampling strategy that selects the part of the time series history that is relevant for forecasting. We achieve this by learning a discrete distribution over relevant time steps by Bayesian optimization. We instantiate this idea with a two-step method that is pre-trained with uniform sampling and then training a lightweight adaptive architecture with adaptive sampling. We show with synthetic and real-world experiments that this method adapts to distribution shift and significantly reduces the forecasting error of the base model for three out of five datasets.
△ Less
Submitted 23 February, 2023;
originally announced February 2023.
-
Bayesian Quantile Matching Estimation
Authors:
Rajbir-Singh Nirwan,
Nils Bertschinger
Abstract:
Due to increased awareness of data protection and corresponding laws many data, especially involving sensitive personal information, are not publicly accessible. Accordingly, many data collecting agencies only release aggregated data, e.g. providing the mean and selected quantiles of population distributions. Yet, research and scientific understanding, e.g. for medical diagnostics or policy advice…
▽ More
Due to increased awareness of data protection and corresponding laws many data, especially involving sensitive personal information, are not publicly accessible. Accordingly, many data collecting agencies only release aggregated data, e.g. providing the mean and selected quantiles of population distributions. Yet, research and scientific understanding, e.g. for medical diagnostics or policy advice, often relies on data access. To overcome this tension, we propose a Bayesian method for learning from quantile information. Being based on order statistics of finite samples our method adequately and correctly reflects the uncertainty of empirical quantiles. After outlining the theory, we apply our method to simulated as well as real world examples. In addition, we provide a python-based package that implements the proposed model.
△ Less
Submitted 11 April, 2022; v1 submitted 14 August, 2020;
originally announced August 2020.
-
Rotation Invariant Householder Parameterization for Bayesian PCA
Authors:
Rajbir S. Nirwan,
Nils Bertschinger
Abstract:
We consider probabilistic PCA and related factor models from a Bayesian perspective. These models are in general not identifiable as the likelihood has a rotational symmetry. This gives rise to complicated posterior distributions with continuous subspaces of equal density and thus hinders efficiency of inference as well as interpretation of obtained parameters. In particular, posterior averages ov…
▽ More
We consider probabilistic PCA and related factor models from a Bayesian perspective. These models are in general not identifiable as the likelihood has a rotational symmetry. This gives rise to complicated posterior distributions with continuous subspaces of equal density and thus hinders efficiency of inference as well as interpretation of obtained parameters. In particular, posterior averages over factor loadings become meaningless and only model predictions are unambiguous. Here, we propose a parameterization based on Householder transformations, which remove the rotational symmetry of the posterior. Furthermore, by relying on results from random matrix theory, we establish the parameter distribution which leaves the model unchanged compared to the original rotationally symmetric formulation. In particular, we avoid the need to compute the Jacobian determinant of the parameter transformation. This allows us to efficiently implement probabilistic PCA in a rotation invariant fashion in any state of the art toolbox. Here, we implemented our model in the probabilistic programming language Stan and illustrate it on several examples.
△ Less
Submitted 12 May, 2019;
originally announced May 2019.
-
Applications of Gaussian Process Latent Variable Models in Finance
Authors:
Rajbir-Singh Nirwan,
Nils Bertschinger
Abstract:
Estimating covariances between financial assets plays an important role in risk management. In practice, when the sample size is small compared to the number of variables, the empirical estimate is known to be very unstable. Here, we propose a novel covariance estimator based on the Gaussian Process Latent Variable Model (GP-LVM). Our estimator can be considered as a non-linear extension of standa…
▽ More
Estimating covariances between financial assets plays an important role in risk management. In practice, when the sample size is small compared to the number of variables, the empirical estimate is known to be very unstable. Here, we propose a novel covariance estimator based on the Gaussian Process Latent Variable Model (GP-LVM). Our estimator can be considered as a non-linear extension of standard factor models with readily interpretable parameters reminiscent of market betas. Furthermore, our Bayesian treatment naturally shrinks the sample covariance matrix towards a more structured matrix given by the prior and thereby systematically reduces estimation errors. Finally, we discuss some financial applications of the GP-LVM.
△ Less
Submitted 18 April, 2019; v1 submitted 8 June, 2018;
originally announced June 2018.
-
Emergent Chiral Spin State in the Mott Phase of a Bosonic Kane-Mele-Hubbard Model
Authors:
Kirill Plekhanov,
Ivana Vasić,
Alexandru Petrescu,
Rajbir Nirwan,
Guillaume Roux,
Walter Hofstetter,
Karyn Le Hur
Abstract:
Recently, the frustrated XY model for spins-1/2 on the honeycomb lattice has attracted a lot of attention in relation with the possibility to realize a chiral spin liquid state. This model is relevant to the physics of some quantum magnets. Using the flexibility of ultra-cold atoms setups, we propose an alternative way to realize this model through the Mott regime of the bosonic Kane-Mele-Hubbard…
▽ More
Recently, the frustrated XY model for spins-1/2 on the honeycomb lattice has attracted a lot of attention in relation with the possibility to realize a chiral spin liquid state. This model is relevant to the physics of some quantum magnets. Using the flexibility of ultra-cold atoms setups, we propose an alternative way to realize this model through the Mott regime of the bosonic Kane-Mele-Hubbard model. The phase diagram of this model is derived using the bosonic dynamical mean-field theory. Focussing on the Mott phase, we investigate its magnetic properties as a function of frustration. We do find an emergent chiral spin state in the intermediate frustration regime. Using exact diagonalization we study more closely the physics of the effective frustrated XY model and the properties of the chiral spin state. This gapped phase displays a chiral order, breaking time-reversal and parity symmetry, but is not topologically ordered (the Chern number is zero).
△ Less
Submitted 10 November, 2017; v1 submitted 21 July, 2017;
originally announced July 2017.
