Massive Data Analysis

Content

Are you ready for “big data” analyses?

“Massive Data Analysis” (MDA) supports your research in statistical databases. It explores "big" time series databases/platforms for any data pattern of your choice to identify strong expected/unexpected correlations automatically. It’s designed e.g. to test/confirm your theses, to identify new correlations and to discover hidden causal relationships (e.g. leading/lagging indicators).

Three steps to new insights:

  1. select the times series of interest (or create your own patterns)
  2. let Massive Data Analysis do the calculation
  3. evaluate the results, identify causal relationships

Methodology

MDA calculates the correlation of numerical patterns against millions of time series by

  • Pearson Coefficients

on a single database platform (here: annual time series for selected AllThatStats databases)

Massive Data Analysis also enables you to shift patterns in time (+/- X periods). It supports research work e.g. for

  • detecting new leading/lagging indicators or to
  • improve analyses when searching for revolving patterns in time.

Results

Further insights can be expected:

  • search for the best series
  • search for the most complete data histories
  • search for alternative dimensions
  • check given time series for completeness and correctness
  • identify poorly documented or unknown time series
  • explore and discover new exciting economic correlations

Massive Data Analysis quantifies the strengths and directions between the variables in question and presents the results (for positive and negative correlations each).

Best matching time series

Massive Data Analysis condenses the results to the best 100 positive and negative hits.

Test Massive Data Analysis right now