Compositional ADAM: An Adaptive Compositional Solver

02/10/2020
by   Rasul Tutunov, et al.
6

In this paper, we present C-ADAM, the first adaptive solver for compositional problems involving a non-linear functional nesting of expected values. We proof that C-ADAM converges to a stationary point in O(δ^-2.25) with δ being a precision parameter. Moreover, we demonstrate the importance of our results by bridging, for the first time, model-agnostic meta-learning (MAML) and compositional optimisation showing fastest known rates for deep network adaptation to-date. Finally, we validate our findings in a set of experiments from portfolio optimisation and meta-learning. Our results manifest significant sample complexity reductions compared to both standard and compositional solvers.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

03/01/2022

Compositional Inverses of AGW-PPs

In this paper, we present two methods to obtain the compositional invers...
10/05/2020

Meta-Learning of Compositional Task Distributions in Humans and Machines

Modern machine learning systems struggle with sample efficiency and are ...
02/12/2020

Distribution-Agnostic Model-Agnostic Meta-Learning

The Model-Agnostic Meta-Learning (MAML) algorithm <cit.> has been celebr...
08/25/2020

Solving Stochastic Compositional Optimization is Nearly as Easy as Solving Stochastic Optimization

Stochastic compositional optimization generalizes classic (non-compositi...
12/15/2020

Are we Forgetting about Compositional Optimisers in Bayesian Optimisation?

Bayesian optimisation presents a sample-efficient methodology for global...
03/19/2021

QROSS: QUBO Relaxation Parameter Optimisation via Learning Solver Surrogates

An increasingly popular method for solving a constrained combinatorial o...
11/03/2020

Novel Compositional Data's Grey Model for Structurally Forecasting Arctic Crude Oil Import

The reserve of crude oil in the Arctic area is abundant. Ice melting is ...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.