arXiv subjects

Koki Hamada

Publications and source records attributed to Koki Hamada.

8 recordsLinked to original sources

arXiv2023

Optimal scheduling in probabilistic imaginary-time evolution on a quantum computer

Ground-state preparation is an important task in quantum computation. The probabilistic imaginary-time evolution (PITE) method is a promising candidate for preparing the ground state of the Hamiltonian, which comprises a single ancilla qubit and forward- and backward-controlled real-time evolution operators. The ground state preparation is a challenging task even in the quantum computation, classified as complexity-class quantum Merlin-Arthur. However, optimal parameters for PITE could potentially enhance the computational efficiency to a certain degree. In this study, we analyze the computational costs of the PITE method for both linear and exponential scheduling of the imaginary-time step size for reducing the computational cost. First, we analytically discuss an error defined as the closeness between the states acted on by exact and approximate imaginary-time evolution operators. The optimal imaginary-time step size and rate of change of imaginary time are also discussed. Subsequently, the analytical discussion is validated using numerical simulations for a one-dimensional Heisenberg chain. From the results, we find that linear scheduling works well in the case of unknown eigenvalues of the Hamiltonian. For a wide range of eigenstates, the linear scheduling returns smaller errors on average. However, the linearity of the scheduling causes problems for some specific energy regions of eigenstates. To avoid these problems, incorporating a certain level of nonlinearity into the scheduling, such as by inclusion of an exponential character, is preferable for reducing the computational costs of the PITE method. The findings of this study can make a significant contribute to the field of ground-state preparation of many-body Hamiltonians on quantum computers.

quant-ph↗

arXiv2021

Efficient decision tree training with new data structure for secure multi-party computation

We propose a secure multi-party computation (MPC) protocol that constructs a secret-shared decision tree for a given secret-shared dataset. The previous MPC-based decision tree training protocol (Abspoel et al. 2021) requires $O(2^hmn\log n)$ comparisons, being exponential in the tree height $h$ and with $n$ and $m$ being the number of rows and that of attributes in the dataset, respectively. The cause of the exponential number of comparisons in $h$ is that the decision tree training algorithm is based on the divide-and-conquer paradigm, where dummy rows are added after each split in order to hide the number of rows in the dataset. We resolve this issue via secure data structure that enables us to compute an aggregate value for every group while hiding the grouping information. By using this data structure, we can train a decision tree without adding dummy rows while hiding the size of the intermediate data. We specifically describes a decision tree training protocol that requires only $O(hmn\log n)$ comparisons when the input attributes are continuous and the output attribute is binary. Note that the order is now \emph{linear} in the tree height $h$. To demonstrate the practicality of our protocol, we implement it in an MPC framework based on a three-party secret sharing scheme. Our implementation results show that our protocol trains a decision tree with a height of 5 in 33 seconds for a dataset of 100,000 rows and 10 attributes.

cs.CR↗

arXiv2021

Designing a Location Trace Anonymization Contest

For a better understanding of anonymization methods for location traces, we have designed and held a location trace anonymization contest that deals with a long trace (400 events per user) and fine-grained locations (1024 regions). In our contest, each team anonymizes her original traces, and then the other teams perform privacy attacks against the anonymized traces. In other words, both defense and attack compete together, which is close to what happens in real life. Prior to our contest, we show that re-identification alone is insufficient as a privacy risk and that trace inference should be added as an additional risk. Specifically, we show an example of anonymization that is perfectly secure against re-identification and is not secure against trace inference. Based on this, our contest evaluates both the re-identification risk and trace inference risk and analyzes their relationship. Through our contest, we show several findings in a situation where both defense and attack compete together. In particular, we show that an anonymization method secure against trace inference is also secure against re-identification under the presence of appropriate pseudonymization. We also report defense and attack algorithms that won first place, and analyze the utility of anonymized traces submitted by teams in various applications such as POI recommendation and geo-data analysis.

cs.CR↗

arXiv2021

Refined Computational Complexities of Hospitals/Residents Problem with Regional Caps

The Hospitals/Residents problem (HR) is a many-to-one matching problem whose solution concept is stability. It is widely used in assignment systems such as assigning medical students (residents) to hospitals. To resolve imbalance in the number of residents assigned to hospitals, an extension called HR with regional caps (HRRC) was introduced. In this problem, a positive integer (called a regional cap) is associated with a subset of hospitals (called a region), and the total number of residents assigned to hospitals in a region must be at most its regional cap. Kamada and Kojima defined strong stability for HRRC and demonstrated that a strongly stable matching does not necessarily exist. Recently, Aziz et al. proved that the problem of determining if a strongly stable matching exists is NP-complete in general. In this paper, we refine Aziz et al.'s result by investigating the computational complexity of the problem in terms of the length of preference lists, the size of regions, and whether or not regions can overlap, and completely classify tractable and intractable cases.

cs.DS↗

arXiv2021

Adam in Private: Secure and Fast Training of Deep Neural Networks with Adaptive Moment Estimation

Privacy-preserving machine learning (PPML) aims at enabling machine learning (ML) algorithms to be used on sensitive data. We contribute to this line of research by proposing a framework that allows efficient and secure evaluation of full-fledged state-of-the-art ML algorithms via secure multi-party computation (MPC). This is in contrast to most prior works, which substitute ML algorithms with approximated "MPC-friendly" variants. A drawback of the latter approach is that fine-tuning of the combined ML and MPC algorithms is required, which might lead to less efficient algorithms or inferior quality ML. This is an issue for secure deep neural networks (DNN) training in particular, as this involves arithmetic algorithms thought to be "MPC-unfriendly", namely, integer division, exponentiation, inversion, and square root. In this work, we propose secure and efficient protocols for the above seemingly MPC-unfriendly computations. Our protocols are three-party protocols in the honest-majority setting, and we propose both passively secure and actively secure with abort variants. A notable feature of our protocols is that they simultaneously provide high accuracy and efficiency. This framework enables us to efficiently and securely compute modern ML algorithms such as Adam and the softmax function "as is", without resorting to approximations. As a result, we obtain secure DNN training that outperforms state-of-the-art three-party systems; our full training is up to 6.7 times faster than just the online phase of the recently proposed FALCON@PETS'21 on a standard benchmark network. We further perform measurements on real-world DNNs, AlexNet and VGG16. The performance of our framework is up to a factor of about 12-14 faster for AlexNet and 46-48 faster for VGG16 to achieve an accuracy of 70% and 75%, respectively, when compared to FALCON.

cs.CR↗

arXiv2020

Strongly Stable and Maximum Weakly Stable Noncrossing Matchings

In IWOCA 2019, Ruangwises and Itoh introduced stable noncrossing matchings, where participants of each side are aligned on each of two parallel lines, and no two matching edges are allowed to cross each other. They defined two stability notions, strongly stable noncrossing matching (SSNM) and weakly stable noncrossing matching (WSNM), depending on the strength of blocking pairs. They proved that a WSNM always exists and presented an $O(n^{2})$-time algorithm to find one for an instance with $n$ men and $n$ women. They also posed open questions of the complexities of determining existence of an SSNM and finding a largest WSNM. In this paper, we show that both problems are solvable in polynomial time. Our algorithms are applicable to extensions where preference lists may include ties, except for one case which we show to be NP-complete. This NP-completeness holds even if each person's preference list is of length at most two and ties appear in only men's preference lists. To complement this intractability, we show that the problem is solvable in polynomial time if the length of preference lists of one side is bounded by one (but that of the other side is unbounded).

cs.DS↗

arXiv2019

Privacy-Preserving Multiple Tensor Factorization for Synthesizing Large-Scale Location Traces with Cluster-Specific Features

With the widespread use of LBSs (Location-based Services), synthesizing location traces plays an increasingly important role in analyzing spatial big data while protecting user privacy. In particular, a synthetic trace that preserves a feature specific to a cluster of users (e.g., those who commute by train, those who go shopping) is important for various geo-data analysis tasks and for providing a synthetic location dataset. Although location synthesizers have been widely studied, existing synthesizers do not provide sufficient utility, privacy, or scalability, hence are not practical for large-scale location traces. To overcome this issue, we propose a novel location synthesizer called PPMTF (Privacy-Preserving Multiple Tensor Factorization). We model various statistical features of the original traces by a transition-count tensor and a visit-count tensor. We factorize these two tensors simultaneously via multiple tensor factorization, and train factor matrices via posterior sampling. Then we synthesize traces from reconstructed tensors, and perform a plausible deniability test for a synthetic trace. We comprehensively evaluate PPMTF using two datasets. Our experimental results show that PPMTF preserves various statistical features including cluster-specific features, protects user privacy, and synthesizes large-scale location traces in practical time. PPMTF also significantly outperforms the state-of-the-art methods in terms of utility and scalability at the same level of privacy.

cs.CR↗

arXiv2019

Strategy-Proof Approximation Algorithms for the Stable Marriage Problem with Ties and Incomplete Lists

In the stable marriage problem (SM), a mechanism that always outputs a stable matching is called a stable mechanism. One of the well-known stable mechanisms is the man-oriented Gale-Shapley algorithm (MGS). MGS has a good property that it is strategy-proof to the men's side, i.e., no man can obtain a better outcome by falsifying a preference list. We call such a mechanism a man-strategy-proof mechanism. Unfortunately, MGS is not a woman-strategy-proof mechanism. Roth has shown that there is no stable mechanism that is simultaneously man-strategy-proof and woman-strategy-proof, which is known as Roth's impossibility theorem. In this paper, we extend these results to the stable marriage problem with ties and incomplete lists (SMTI). Since SMTI is an extension of SM, Roth's impossibility theorem takes over to SMTI. Therefore, we focus on the one-sided-strategy-proofness. In SMTI, one instance can have stable matchings of different sizes, and it is natural to consider the problem of finding a largest stable matching, known as MAX SMTI. Thus we incorporate the notion of approximation ratio used in the theory of approximation algorithms. We say that a stable-mechanism is $c$-approximate-stable mechanism if it always returns a stable matching of size at least $1/c$ of a largest one. We also consider a restricted variant of MAX SMTI, which we call MAX SMTI-1TM, where only men's lists can contain ties. Our results are summarized as follows: (i) MAX SMTI admits both a man-strategy-proof 2-approximate-stable mechanism and a woman-strategy-proof 2-approximate-stable mechanism. (ii) MAX SMTI-1TM admits a woman-strategy-proof 2-approximate-stable mechanism. (iii) MAX SMTI-1TM admits a man-strategy-proof 1.5-approximate-stable mechanism. All these results are tight in terms of approximation ratios. Also, all these strategy-proofness results apply for strategy-proofness against coalitions.

cs.GT↗