Sandbox‎ > ‎Archive‎ > ‎IPT 2008-09‎ > ‎Ria's Page‎ > ‎RIA CLASSWORK‎ > ‎

DSS Types

posted Jun 10, 2009, 5:15 PM by Unknown user   [ updated Jun 11, 2009, 6:19 AM by Eddie Woo ]
  1. Spreadsheets
      (a) Describe the kinds of analysis supported by a spreadsheet.
Types of analysis that can be supported by a spreadsheet are primarily numerical as it normally analyses and organises figures that can be manually inserted or scanned. Numerical data is normally calculated and manipulated to assist in concluding with a decision. 
What if analysis is also supported by a spreadsheet as it can change specific variables in order to fixate a result that may not be realistic or simulated.
Goal seeking analysis is changing faactors in accordance wiith a set goal that you have instructed. A spreadsheet can allow you to view the needed factors to seek the goal.
 
 (b) Identify one example each of situations/decisions that would be supported (i) effectively by a spreadsheet or (ii) ineffectively by a spreadsheet.
 
i) keeping a weekly record of materials that are sold within a small shop, a weekly record will keep the spreadsheet limited in size and comparisons can be seen on a monthly basis or every 3 months. This can help view patterns nad trends and assist the owners with making future decisions.
ii) a spreadsheet would not be compatible with a circumstance that may include a person choosing what clothes to wear on a specific day from their extensive wardrobe. Numarical data does not take into account personal choices and personal factors, therefore a spreadsheet whould not be useful in this circumstance.



 

  1. Databases
      Explain how a relational database can support decision-making in a way that a spreadsheet cannot.

Relational databases allow multiple data sources to be able to be analysed and drawn upon when creating a solution, as opposed to a spreadsheet. The complexity of such a database as opposed to that of a spreadsheet allows extensive speed when creating a solution as well as easier access, all data being stored on the one file instead of many spreadsheets scattered on the desktop. Looking for data that is also specific to your problem is accessed with more ease in a realtional database through the use of SQL queries, this is not the case with spreadsheets as they are not normally as extensive and all the data is present in the one file. Data in a relational database is also linked, with similar information being found quicker than manually reading every record on a spreadsheet.

  1. Expert Systems
      (a) Identify the aspects of human intelligence that are imitated by artificial intelligence (AI).
      (b) Outline some of the difficulties in designing and implementing an AI-based DSS.
      (c) Carefully read the infobox on p193. For the four examples provided, describe the benefits and limitations of the AI-based DSS in each case.
      (d) Account for why expert systems would be designed to be domain-specific.

  2. Neural Networks
       (a) Identify the basic concept that neural networks are based upon. Propose arguments for and against the use of this particular concept as the model of a computer-based information system.
       (b) Neural networks and expert systems are both designed by human software engineers. Contrast the relationships between these two decision support systems and their designers.
       (c) A neural network's designer is often incapable of providing an explanation for the decisions made by the network as a whole. Explain how this can be the case.
       (d) Compare the characteristics of parallel processing systems and neural networks
Comments