-
Schizophrenia-mimicking layers outperform conventional neural network layers
Authors:
Ryuta Mizutani,
Senta Noguchi,
Rino Saiga,
Yuichi Yamashita,
Mitsuhiro Miyashita,
Makoto Arai,
Masanari Itokawa
Abstract:
We have reported nanometer-scale three-dimensional studies of brain networks of schizophrenia cases and found that their neurites are thin and tortuous compared to healthy controls. This suggests that connections between distal neurons are suppressed in microcircuits of schizophrenia cases. In this study, we applied these biological findings to the design of schizophrenia-mimicking artificial neur…
▽ More
We have reported nanometer-scale three-dimensional studies of brain networks of schizophrenia cases and found that their neurites are thin and tortuous compared to healthy controls. This suggests that connections between distal neurons are suppressed in microcircuits of schizophrenia cases. In this study, we applied these biological findings to the design of schizophrenia-mimicking artificial neural network to simulate the observed connection alteration in the disorder. Neural networks having a "schizophrenia connection layer" in place of a fully connected layer were subjected to image classification tasks using the MNIST and CIFAR-10 datasets. The results revealed that the schizophrenia connection layer is tolerant to overfitting and outperforms a fully connected layer. The outperformance was observed only for networks using band matrices as weight windows, indicating that the shape of the weight matrix is relevant to the network performance. A schizophrenia convolution layer was also tested using the VGG configuration, showing that 60% of the kernel weights of the last three convolution layers can be eliminated without loss of accuracy. The schizophrenia layers can be used instead of conventional layers without any change in the network configuration and training procedures; hence, neural networks can easily take advantage of these layers. The results of this study suggest that the connection alteration found in schizophrenia is not a burden to the brain, but has functional roles in brain performance.
△ Less
Submitted 1 April, 2022; v1 submitted 22 September, 2020;
originally announced September 2020.
-
A Study on Optimal Beam Patterns for Single User Massive MIMO Transmissions
Authors:
Maki Arai,
Kei Sakaguchi,
Kiyomichi Araki
Abstract:
This paper proposes optimal beam patterns of analog beamforming for SU (Single User) massive MIMO (Multi-Input Multi-Output) transmission systems. For hybrid beamforming in SU massive MIMO systems, there are several design parameters such as beam patterns, the number of beams (streams), the shape of array antennas, and so on. In conventional hybrid beamforming, rectangular patch array antennas imp…
▽ More
This paper proposes optimal beam patterns of analog beamforming for SU (Single User) massive MIMO (Multi-Input Multi-Output) transmission systems. For hybrid beamforming in SU massive MIMO systems, there are several design parameters such as beam patterns, the number of beams (streams), the shape of array antennas, and so on. In conventional hybrid beamforming, rectangular patch array antennas implemented on a planar surface with linear phase shift beam patterns have been used widely. However, it remains unclear whether existing configurations are optimal or not. Therefore, we propose a method using OBPB (Optimal Beam Projection Beamforming) for designing configuration parameters of the hybrid beamforming. By using the method, the optimal beam patterns are derived first, and are projected on the assumed surface to calculate the achievable number of streams and the resulting channel capacity. The results indicate OBPB with a spherical surface yields at least 3.5 times higher channel capacity than conventional configurations.
△ Less
Submitted 14 December, 2018;
originally announced December 2018.
-
Optimal Design Method of MIMO Antenna Directivities and Corresponding Current Distributions by Using Spherical Mode Expansion
Authors:
Maki Arai,
Masashi Iwabuchi,
Kei Sakaguchi,
Kiyomichi Araki
Abstract:
This paper proposes a new methodology to design optimal antennas for MIMO (Multi-Input Multi-Output) communication systems by using spherical mode expansion. Given spatial channel properties of a MIMO channel, such as the angular profile at both sides, the optimal MIMO antennas should provide the largest channel capacity with a constraint of the limited implementation space (volume). In designing…
▽ More
This paper proposes a new methodology to design optimal antennas for MIMO (Multi-Input Multi-Output) communication systems by using spherical mode expansion. Given spatial channel properties of a MIMO channel, such as the angular profile at both sides, the optimal MIMO antennas should provide the largest channel capacity with a constraint of the limited implementation space (volume). In designing a conventional MIMO antenna, first the antenna structure (current distribution) is determined, second antenna directivity is calculated based on the current distribution, and thirdly MIMO channel capacity is calculated by using given angular profiles and obtained antenna directivity. This process is repeated by adjusting the antenna structure until the performance satisfies a predefined threshold. To the contrary, this paper solves the optimization problem analytically and finally gives near optimal antenna structure (current distribution) without any greedy search. In the proposed process, first the optimal directivity of MIMO antennas is derived by applying spherical mode expansion to the angular profiles, and second a far-near field conversion is applied on the derived optimal directivity to achieve near optimal current distributions on a limited surface. The effectiveness of the proposed design methodology is validated via numerical calculation of MIMO channel capacity as in the conventional design method while giving near optimal current distribution with constraint of an antenna structure derived from proposed methodology.
△ Less
Submitted 5 June, 2017;
originally announced June 2017.
-
Estimating the resolution of real images
Authors:
Ryuta Mizutani,
Rino Saiga,
Susumu Takekoshi,
Chie Inomoto,
Naoya Nakamura,
Makoto Arai,
Kenichi Oshima,
Masanari Itokawa,
Akihisa Takeuchi,
Kentaro Uesugi,
Yasuko Terada,
Yoshio Suzuki
Abstract:
Image resolvability is the primary concern in imaging. This paper reports an estimation of the full width at half maximum of the point spread function from a Fourier domain plot of real sample images by neither using test objects, nor defining a threshold criterion. We suggest that this method can be applied to any type of image, independently of the imaging modality.
Image resolvability is the primary concern in imaging. This paper reports an estimation of the full width at half maximum of the point spread function from a Fourier domain plot of real sample images by neither using test objects, nor defining a threshold criterion. We suggest that this method can be applied to any type of image, independently of the imaging modality.
△ Less
Submitted 2 March, 2017;
originally announced March 2017.
-
Neural Feature Embedding for User Response Prediction in Real-Time Bidding (RTB)
Authors:
Enno Shioji,
Masayuki Arai
Abstract:
In the area of ad-targeting, predicting user responses is essential for many applications such as Real-Time Bidding (RTB). Many of the features available in this domain are sparse categorical features. This presents a challenge especially when the user responses to be predicted are rare, because each feature will only have very few positive examples. Recently, neural embedding techniques such as w…
▽ More
In the area of ad-targeting, predicting user responses is essential for many applications such as Real-Time Bidding (RTB). Many of the features available in this domain are sparse categorical features. This presents a challenge especially when the user responses to be predicted are rare, because each feature will only have very few positive examples. Recently, neural embedding techniques such as word2vec which learn distributed representations of words using occurrence statistics in the corpus have been shown to be effective in many Natural Language Processing tasks. In this paper, we use real-world data set to show that a similar technique can be used to learn distributed representations of features from users' web history, and that such representations can be used to improve the accuracy of commonly used models for predicting rare user responses.
△ Less
Submitted 18 May, 2017; v1 submitted 2 February, 2017;
originally announced February 2017.
