Page 138 - Data Science class 11
P. 138

3.1.6 What can and cannot be forecasted?

        It is possible to predict the exact moment of sunrise tomorrow morning. The lottery numbers for tomorrow, on the
        other hand, cannot be predicted with any confidence.

        The predictability of an event or a quantity depends on many factors like:
           • Our understanding of the factors that contribute to it;

           • Volume of data available;
           • Whether the forecasts can have an impact on the things we are trying to forecast.
        If we consider forecasts of electricity demand, they can be highly accurate because all three conditions mentioned
        above are usually satisfied. Here, we can easily identify the contributing factors: electricity demand is driven majorly
        by temperatures, with smaller effects for calendar variations such as holidays, and economic conditions. Forecasts
        may be extremely accurate if there is enough historical data on energy demand and weather conditions, and we have
        the skills to create a strong model that links electricity demand and the primary driver variables.
        Whereas, when forecasting currency exchange rates, only one of the conditions holds true: there is an excess of
        available data. We do, however, have a limited understanding of the elements that influence exchange rates, and
        exchange rate forecasts have a direct impact on the rates themselves.


        3.2 anaLYtic stUDies

        An analytical study is one in which action will be taken on a cause system to enhance the performance of the system
        of interest in the future. It focuses on prediction or forecasting.

        The analytical studies can be of two types: Experimental studies or Observational studies. We will mainly learn about
        observational studies.

                                                     Analytic Studies
                                            Examine etiology and causal associations



                                Experimental Studies                  Observational Studies
                             One or more factors altered and           Observations made
                                   effects examined                  without any interventions


        3.2.1 experimental studies
        In a controlled experiment, we assign people or things to groups and apply some treatment to one of the groups,
        while the other group does not receive the treatment.

        3.2.2 Observational studies

        An observational study can be defined as a procedure in which the subjects are just observed and the results are then
        noted. During the investigation, nobody tries to interfere with the subject to affect the outcome.

        In an observational study, we measure or survey members of a sample without trying to affect them.

        Principle of Observational Studies
        According to Bandura, learning can occur by watching others and then modeling what they do or say. This is known
        as 'observational learning.' There are specific steps in the process of modeling that must be followed, if learning is to
        be successful. These steps include attention, retention, reproduction, and motivation.

        Examples of Observational Study
        Let us have an insight into different examples of observational study to have a better understanding on this topic.

          136   Touchpad Data Science-XI
   133   134   135   136   137   138   139   140   141   142   143