no code implementations • 16 Dec 2023 • Tatsuya Akutsu, Avraham A. Melkman, Atsuhiro Takasu
We also show that a bag of $n$ decision trees can be represented by a bag of $T$ decision trees each with polynomial size if $n-T$ is a constant and a small classification error is allowed.
no code implementations • 27 Apr 2023 • Jianshen Zhu, Naveed Ahmed Azam, Kazuya Haraguchi, Liang Zhao, Hiroshi Nagamochi, Tatsuya Akutsu
A novel framework for designing the molecular structure of chemical compounds with a desired chemical property has recently been proposed.
1 code implementation • 13 Sep 2022 • Jianshen Zhu, Naveed Ahmed Azam, Shengjuan Cao, Ryota Ido, Kazuya Haraguchi, Liang Zhao, Hiroshi Nagamochi, Tatsuya Akutsu
A set of graph theoretical descriptors in the feature function plays a key role to derive a compact formulation of such an MILP.
no code implementations • 21 Dec 2021 • Tatsuya Akutsu, Avraham A. Melkman
In this paper, we study the size and width of autoencoders consisting of Boolean threshold functions, where an autoencoder is a layered neural network whose structure can be viewed as consisting of an encoder, which compresses an input vector to a lower dimensional vector, and a decoder which transforms the low-dimensional vector back to the original input vector exactly (or approximately).
no code implementations • 24 Aug 2021 • Ryota Ido, Shengjuan Cao, Jianshen Zhu, Naveed Ahmed Azam, Kazuya Haraguchi, Liang Zhao, Hiroshi Nagamochi, Tatsuya Akutsu
For this, we introduce a new way of representing a polymer as a form of monomer and define new descriptors that feature the structure of polymers.
no code implementations • 23 Aug 2021 • Naveed Ahmed Azam, Jianshen Zhu, Kazuya Haraguchi, Liang Zhao, Hiroshi Nagamochi, Tatsuya Akutsu
In the framework, a chemical graph with a target chemical value is inferred as a feasible solution of a mixed integer linear program that represents a prediction function and other requirements on the structure of graphs.
1 code implementation • 6 Jul 2021 • Jianshen Zhu, Naveed Ahmed Azam, Kazuya Haraguchi, Liang Zhao, Hiroshi Nagamochi, Tatsuya Akutsu
In the framework, we first define a feature vector $f(C)$ of a chemical graph $C$ and construct an ANN that maps $x=f(C)$ to a predicted value $\eta(x)$ of a chemical property $\pi$ to $C$.
1 code implementation • 21 Sep 2020 • Naveed Ahmed Azam, Jianshen Zhu, Yanming Sun, Yu Shi, Aleksandar Shurbevski, Liang Zhao, Hiroshi Nagamochi, Tatsuya Akutsu
In the second phase, given a target value $y^*$ of property $\pi$, a feature vector $x^*$ is inferred by solving an MILP formulated from the trained ANN so that $\psi(x^*)$ is close to $y^*$ and then a set of chemical structures $G^*$ such that $f(G^*)= x^*$ is enumerated by a graph search algorithm.
Data Structures and Algorithms Computational Engineering, Finance, and Science 05C92, 92E10, 05C30, 68T07, 90C11, 92-04
no code implementations • 21 Apr 2020 • Avraham A. Melkman, Sini Guo, Wai-Ki Ching, Pengyu Liu, Tatsuya Akutsu
An autoencoder is a layered neural network whose structure can be viewed as consisting of an encoder, which compresses an input vector of dimension $D$ to a vector of low dimension $d$, and a decoder which transforms the low-dimensional vector back to the original input vector (or one that is very similar).
no code implementations • 3 Jun 2014 • Hitoshi Koyano, Morihiro Hayashida, Tatsuya Akutsu
However, this non-one-to-one conversion involves a loss of information and makes it impossible to evaluate, using probability theory, the generalization error of a learning machine, considering that the given data to train and test the machine are strings generated according to probability laws.