主催　TEDS (TRANSDISCIPLINARY ECONOMETRICS & DATA SCIENCE SEMINAR)
https://teds-datascience.github.io/seminars/
共催　共同利用・共同研究拠点プロジェクト研究 (Joint Research Program of KIER, Kyoto University)
共催　計量経済学セミナー(Econometrics Seminar of KIER, Kyoto University)
（登録者のみ参加可能）

要旨：We propose a scalable regression model with spatially and temporally varying co- efficients based on Moran’s eigenvectors and efficient computation algorithms. Regression models that consider spatiotemporal non-stationarity are important because many real-world datasets, such as housing prices, are tied to geographical and tempo- ral locations. Although geographically weighted regression (GWR) and its variants are widely used to model spatially varying coefficients, they cannot handle large datasets. We employ an alternative modelling method of spatially varying coefficients based on Moran’s eigenvectors and extend it to handle large spatiotemporal datasets. Additionally, we introduce a scalable learning algorithm that exploits the model structures based on the Kalman filter and the expectation—maximisation algorithm. Our scalable algorithm is efficient even for large datasets that cannot be handled by GWR. To evaluate the performance of the proposed model, we applied it to a housing market dataset collected in Tokyo, Japan. The results show that the predictive performance of the proposed model is comparable to that of GWR while increasing the computational speed. Moreover, larger datasets can accelerate the algorithm convergence.

主催　TEDS (TRANSDISCIPLINARY ECONOMETRICS & DATA SCIENCE SEMINAR)
https://teds-datascience.github.io/seminars/
共催　共同利用・共同研究拠点プロジェクト研究 (Joint Research Program of KIER, Kyoto University)
共催　計量経済学セミナー(Econometrics Seminar of KIER, Kyoto University)

（登録者のみ参加可能）　

要旨： We study the impact of telecommuting in a monocentric city which produces (i) a tradable consumption good using skilled and unskilled labor and (ii) a non-tradable consumer service provided by unskilled workers at the city center to the skilled workers. Commuting costs are proportional to wages. When the WFH share is low, the skilled reside near the CBD and all workers earn more under WFH. By contrast, a high WFH share lowers both wages and leads the skilled to reside in the suburbs. Telecommuting leads to lower urban costs in the latter case, but not in the former. We then consider two cities that have different productivities. WFH allows skilled workers of the more productive city to reside in the less productive city where housing is cheaper while keeping their job in the more productive city. The flow of this type of inter-city commuters first increases and, then, decreases with the WFH share. Likewise, skilled workers of the less productive city may take a job in the more productive city while keeping their residence in the less productive city. The flow of the second type of inter-city commuters increases with the WFH share. For these commuting patterns to arise, the two employment centers must be connected by a link that allows workers to travel at relatively low costs.

要旨：COVID-19が猛威を振るう昨今、感染拡大に寄与した要因を明らかにすることは喫緊の課題となっている。陽性者数や死者数の分析には、ポアソン回帰やその拡張手法が用いられてきた。しかしながら、ポアソン回帰には次の課題が残されている：(i)ゼロ値の多いカウントデータの場合は解が識別できず、この問題はモデルが複雑化するほど深刻化する；(ii)そもそも陽性者数や死者数がポアソン分布に従うとは限らない。そこで本研究では、課題(i)、(ii)に対処することで、多種多様な要因の影響を安定的に推定するための新たな回帰手法を開発する。同開発では、課題(i)に対処するための新たな対数線形近似を提案し、次に課題(ii)に対処するためのデータ分布の自動推定手法を提案する。最後に、開発した手法をCOVID-19の地理的要因の分析に応用することで、人流、世代、政策などの各種要因の感染拡大に対する影響を評価する。