The regularized least squares algorithm and the problem of learning halfspaces

Minh, Hà Quang

doi:10.1016/j.ipl.2011.01.011

Cited by 3 publications

(1 citation statement)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where α i , β i and γ i are determined based on the leastsquare method that is usually used for forecast work [14]. With three crucial parameters α 1 , β 1 and γ 1 determined, we obtain the first sine function S 1 (t) as well as the first decomposed sequence (or say, the first yielded sequence) Y t1 .…”

Section: A Msfd Methodsmentioning

confidence: 99%

Warming substantiated by multiple sine functions decomposition of multiple cities' temperature data

Zhang

Qiu

et al. 2015

2015 Chinese Automation Congress (CAC)

View full text Add to dashboard Cite

In recent years, extensive efforts have been committed to the temperature forecast work, and more and more studies have indicated that the temperature is rising. For the purpose of forecasting long-term temperature trend, a temperature forecast method named multiple sine functions decomposition (MSFD) method is presented in this paper. Based on the numerical experiments for 12 Asia-Pacific (APAC) cities, the efficacy of the MSFD method and the warming trend of APAC region are substantiated. The core concept of the MSFD method is that, by decomposing a historical temperature sequence into sine waveforms, multiple sine functions are thus obtained and applied to forecasting the future temperature trend. Different degrees of warming in the ensuing 200 years are forecasted from the results of numerical experiments. In addition, we find that the high-latitude cities are generally more evident in terms of temperature change compared with those low-latitude cities. In summary, with the most possibility, the general temperature trend of APAC region is upward.

show abstract

Section: A Msfd Methodsmentioning

confidence: 99%