Fully-Dynamic Approximate Decision Trees With Worst-Case Update Time Guarantees

• URL: http://arxiv.org/abs/2302.03994v2
• Date: Fri, 10 Feb 2023 09:14:35 GMT
• Title: Fully-Dynamic Approximate Decision Trees With Worst-Case Update Time Guarantees
• Authors: Marco Bressan and Mauro Sozio
• Abstract summary: We give the first algorithm that maintains an approximate decision tree over an arbitrary sequence of insertions and deletions of labeled examples. We provide a deterministic algorithm that maintains an $epsilon$-approximate tree using $O!left(fracd, f(n)n operatornamepolyfrachepsilonright)$ operations per update.
• Score: 3.5509551353363644
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We give the first algorithm that maintains an approximate decision tree over an arbitrary sequence of insertions and deletions of labeled examples, with strong guarantees on the worst-case running time per update request. For instance, we show how to maintain a decision tree where every vertex has Gini gain within an additive $\alpha$ of the optimum by performing $O\Big(\frac{d\,(\log n)^4}{\alpha^3}\Big)$ elementary operations per update, where $d$ is the number of features and $n$ the maximum size of the active set (the net result of the update requests). We give similar bounds for the information gain and the variance gain. In fact, all these bounds are corollaries of a more general result, stated in terms of decision rules -- functions that, given a set $S$ of labeled examples, decide whether to split $S$ or predict a label. Decision rules give a unified view of greedy decision tree algorithms regardless of the example and label domains, and lead to a general notion of $\epsilon$-approximate decision trees that, for natural decision rules such as those used by ID3 or C4.5, implies the gain approximation guarantees above. The heart of our work provides a deterministic algorithm that, given any decision rule and any $\epsilon > 0$, maintains an $\epsilon$-approximate tree using $O\!\left(\frac{d\, f(n)}{n} \operatorname{poly}\frac{h}{\epsilon}\right)$ operations per update, where $f(n)$ is the complexity of evaluating the rule over a set of $n$ examples and $h$ is the maximum height of the maintained tree.

Related papers

• Tight Bounds for $\gamma$-Regret via the Decision-Estimation Coefficient [88.86699022151598]
  We give a statistical characterization of the $gamma$-regret for arbitrary structured bandit problems. The $gamma$-regret emerges in structured bandit problems over a function class $mathcalF$ where finding an exact optimum of $f in mathcalF$ is intractable.
  arXiv  Detail & Related papers  (2023-03-06T17:54:33Z)
• Fully-Dynamic Decision Trees [3.0058005235097114]
  We develop the first fully dynamic algorithm that maintains a decision tree over an arbitrary sequence of insertions and deletions of labeled examples. For real-valued features the algorithm has an amortized running time per insertion/deletion of $Obig(fracd log3 nepsilon2big)$. We show that any algorithm with similar guarantees uses amortized running time $Omega(d)$ and space $tildeOmega (n d)$.
  arXiv  Detail & Related papers  (2022-12-01T18:58:19Z)
• Superpolynomial Lower Bounds for Decision Tree Learning and Testing [5.117810469621253]
  We prove, under randomized ETH, superpolynomial lower bounds for two basic problems. Our results imply new lower bounds for distribution-free PAC learning and testing of decision trees. We show our lower bound for learning decision trees can be improved to $nOmega(log s)$.
  arXiv  Detail & Related papers  (2022-10-12T16:25:48Z)
• Corralling a Larger Band of Bandits: A Case Study on Switching Regret for Linear Bandits [99.86860277006318]
  We consider the problem of combining and learning over a set of adversarial algorithms with the goal of adaptively tracking the best one on the fly. The CORRAL of Agarwal et al. achieves this goal with a regret overhead of order $widetildeO(sqrtd S T)$ where $M$ is the number of base algorithms and $T$ is the time horizon. Motivated by this issue, we propose a new recipe to corral a larger band of bandit algorithms whose regret overhead has only emphlogarithmic dependence on $M$ as long
  arXiv  Detail & Related papers  (2022-02-12T21:55:44Z)
• Private Stochastic Convex Optimization: Optimal Rates in $\ell_1$ Geometry [69.24618367447101]
  Up to logarithmic factors the optimal excess population loss of any $(varepsilon,delta)$-differently private is $sqrtlog(d)/n + sqrtd/varepsilon n.$ We show that when the loss functions satisfy additional smoothness assumptions, the excess loss is upper bounded (up to logarithmic factors) by $sqrtlog(d)/n + (log(d)/varepsilon n)2/3.
  arXiv  Detail & Related papers  (2021-03-02T06:53:44Z)
• Testing and reconstruction via decision trees [19.304587350775385]
  We study sublinear and local computation algorithms for decision trees, focusing on testing and reconstruction. A tester that runs in $mathrmpoly(log s, 1/varepsilon)cdot nlog n$ time makes $mathrmpoly(log s,1/varepsilon)cdot n$ queries to an unknown function.
  arXiv  Detail & Related papers  (2020-12-16T04:18:00Z)
• Universal guarantees for decision tree induction via a higher-order splitting criterion [16.832966312395126]
  Our algorithm achieves provable guarantees for all target functions $f: -1,1n to -1,1$ with respect to the uniform distribution. The crux of our extension is a new splitting criterion that takes into account the correlations between $f$ and small subsets of its attributes. Our algorithm satisfies the following guarantee: for all target functions $f : -1,1n to -1,1$, sizes $sin mathbbN$, and error parameters $epsilon$, it constructs a decision
  arXiv  Detail & Related papers  (2020-10-16T21:20:45Z)
• An Optimal Separation of Randomized and Quantum Query Complexity [67.19751155411075]
  We prove that for every decision tree, the absolute values of the Fourier coefficients of a given order $ellsqrtbinomdell (1+log n)ell-1,$ sum to at most $cellsqrtbinomdell (1+log n)ell-1,$ where $n$ is the number of variables, $d$ is the tree depth, and $c>0$ is an absolute constant.
  arXiv  Detail & Related papers  (2020-08-24T06:50:57Z)
• Revisiting Modified Greedy Algorithm for Monotone Submodular Maximization with a Knapsack Constraint [75.85952446237599]
  We show that a modified greedy algorithm can achieve an approximation factor of $0.305$. We derive a data-dependent upper bound on the optimum. It can also be used to significantly improve the efficiency of such algorithms as branch and bound.
  arXiv  Detail & Related papers  (2020-08-12T15:40:21Z)
• Model-Free Reinforcement Learning: from Clipped Pseudo-Regret to Sample Complexity [59.34067736545355]
  Given an MDP with $S$ states, $A$ actions, the discount factor $gamma in (0,1)$, and an approximation threshold $epsilon > 0$, we provide a model-free algorithm to learn an $epsilon$-optimal policy. For small enough $epsilon$, we show an improved algorithm with sample complexity.
  arXiv  Detail & Related papers  (2020-06-06T13:34:41Z)
• Provable guarantees for decision tree induction: the agnostic setting [16.784355746717562]
  We give strengthened provable guarantees on the performance of widely employed and empirically successful sl top-down decision tree learnings. We show that for all monotone functions$f$ and parameters $sin mathN$, theses construct a decision tree of size $stildeO((log s)/varepsilon2)$ that achieves error. We complement our algorithmic guarantee with a near-matching $stildeOmega(log s)$ lower bound.
  arXiv  Detail & Related papers  (2020-06-01T06:44:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.