First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper proposes a new regularization method for structured prediction. The idea is relatively straightforward: a linear chain model is segmented into smaller subchains, each of which is added as an independent training example. Theorems are provided (with proofs in the supplement) showing how this regularization can reduce generalization risk and accelerate convergence rates. Empirical comparisons with state of the art approaches suggest that the resulting method is both faster and more accurate.

We thank all the reviewers for the time and expertise invested in these reviews. Q: What is the meaning of every notation? Their corresponding lowercase letter refer to one instance in the set, e.g. Q: What is the relationship to other Transfer Learning/Imitation Learning method? Since there are no major flaws pointed out in the review, could the reviewer please raise the overall score?

Neural Information Processing Systems http://nips.cc/

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. The paper presents a new recursive neural network architecture for semantic scene labeling and shows that it outperforms previous approaches on two standard datasets in terms of pixel accuracy. The paper is generally very well written and the proposed model seems quite natural and conceptually clean compared to its main competitors. My only major concern is that the paper doesn't separately evaluate the effects of the combiner and decombiner networks. An even simpler model could use the combiner network to recursively collapse everything to a single root node (as is done already) but then directly feed the output of the F_sem network along with the root node F_com features into each corresponding F_lab network.

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. Summary: This paper proposes boosting algorithms and analyze them from an online learning perspective. First, they propose a boosting algorithm based on the update of the online mirror descent(MABoost). Then they show a smooth version of MABoost, i.e., a variant of MABoost which creates only smooth distributions over examples. Further, they propose sparse or lazy versions of MABoost and show their convergence proofs.

These certificates map out-of-distribution examples to non-zero values, signaling epistemic uncertainty.

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. This paper proposes a novel approach to human pose estimation, consisting of a deep convolutional network for part detection and a higher-level spatial model that is motivated as a graphical model, but actually incorporated into the overall deep network as a particular sub-net that has the plausible interpretation of performing a single round of message passing. The system is trained in three steps. In the first two steps, the deep convolutional part detector and the spatial model are trained individually (the spatial message passing network uses the heat map output of the part detector), while in the third step, the unified network is jointly trained via back propagation. Even though the convolutional part detector alone is already a state-of-the-art system, the spatial model is shown to improve results considerably, with even further improvements gained via the joint training procedure.

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. The paper presents a simple novel model for structured sparsity in spike-and-slab models and an expectation propagation algorithm for Bayesian inference. It is written clearly and accompanied by interesting examples and comparisons with other approaches. Q2: Please summarize your review in 1-2 sentences Although there are some previous works in this area, this particular approach is simple and novel. First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance.

This paper introduces Deep Statistical Solvers (DSS), a new class of trainable solvers for optimization problems, arising e.g., from system simulations. The key idea is to learn a solver that generalizes to a given distribution of problem instances. This is achieved by directly using as loss the objective function of the problem, as opposed to most previous Machine Learning based approaches, which mimic the solutions attained by an existing solver.

This paper examines replication portfolio construction in incomplete markets - a key problem in financial engineering with applications in pricing, hedging, balance sheet management, and energy storage planning. We model this as a two-player game between an investor and the market, where the investor makes strategic bets on future states while the market reveals outcomes. Inspired by the success of Monte Carlo Tree Search in stochastic games, we introduce an AlphaZero-based system and compare its performance to deep hedging - a widely used industry method based on gradient descent. Through theoretical analysis and experiments, we show that deep hedging struggles in environments where the $Q$-function is not subject to convexity constraints - such as those involving non-convex transaction costs, capital constraints, or regulatory limitations - converging to local optima. We construct specific market environments to highlight these limitations and demonstrate that AlphaZero consistently finds near-optimal replication strategies. On the theoretical side, we establish a connection between deep hedging and convex optimization, suggesting that its effectiveness is contingent on convexity assumptions. Our experiments further suggest that AlphaZero is more sample-efficient - an important advantage in data-scarce, overfitting-prone derivative markets.