id author title date pages extension mime words sentences flesch summary cache txt work_kqsqmw4gcnewjmjpwwpl7gzafa Jari Avikainen A Method for Wavelet-Based Time Series Analysis of Historical Newspapers 2019 50 .pdf application/pdf 12297 968 71 A Method for Wavelet-Based Time Series Analysis of Historical Newspapers historical newspapers, wavelet transform, step detection for change detection in time series generated from historical newspaper data. of the transformed time series are used for step detection similar to the analysis by Sadler As an example of a (partial) wavelet transform using real data, Figure 3.2 shows a time Figure 3.2: A time series of newspaper coverage per month (a.) and the first three scales of the wavelet sharp edges produce maxima across multiple scales in the wavelet transformation at approximately same location, and therefore the multiscale products will have large values at Figure 3.5: The proportion of assumed data used for calculating the wavelet product of a time series of Figure 4.3.a. shows the time series used to test the effect of step width to the corresponding Figure 4.3: The signals used for testing the effects of step width (a.) and height (b.) to the wavelet ./cache/work_kqsqmw4gcnewjmjpwwpl7gzafa.pdf ./txt/work_kqsqmw4gcnewjmjpwwpl7gzafa.txt