{"id":60,"date":"2020-01-21T11:26:15","date_gmt":"2020-01-21T11:26:15","guid":{"rendered":"http:\/\/biosis.tu.ac.kr\/?p=60"},"modified":"2020-01-21T11:26:15","modified_gmt":"2020-01-21T11:26:15","slug":"the-similarity-between-two-time-series-signals","status":"publish","type":"post","link":"https:\/\/pkai.pknu.ac.kr\/index.php\/2020\/01\/21\/the-similarity-between-two-time-series-signals\/","title":{"rendered":"The similarity between two-time series signals"},"content":{"rendered":"\n
\"\"<\/figure>
\n

Calculating similarity of two time-series signal is not an easy issue. A naive way is directly comparing pointwise after sampling, but it often is not a good idea because it results in completely different results to human perception. <\/p>\n<\/div><\/div>\n\n\n

\"\"<\/figure>
\n

A typical human would match points in consideration of its shape and curvature. We may need algorithms working like a human being. <\/p>\n<\/div><\/div>\n\n\n

\"\"<\/figure>
\n

We proposed a novel algorithm for this issue, by extending DTW (dynamic time warping), named Dyamic Positional Warping (DPW)<\/em><\/strong>. We also found that DPW outperforms in the recognition tasks of other problems such as EEG. <\/p>\n<\/div><\/div>\n\n\n

Paper lists<\/p>\n\n\n