Lit Review

ABE is good for prediction especially when different data types are used.


Build clusters. Assign new project to one of the aforementioned clusters


Effort Estimation Using Analogy
Shepperd, Schofield, Kitchenham : ”Effort Estimation Using Analogy” , Proceedings ICSE.18 (1996) 170-178


When compared with regression and step-wise estimates, analogy performed with superior accuracy, and had the advantage of being self calibrating. In being an accurate predictor, estimation by analogy should become one of the techniques considered for estimation in the early stage of development. When using several different estimation techniques, the variance in predictions can also be used as a risk indicator.


1. How best to describe projects
2. Measure similarity, and how much confidence
3. How do we use known effort values to drive estimate (means/weighted means/etc.


Differing datasets
  1. Number of observations
  2. Number of variables
Effect software effort estimation DEMANDS more than one technique


I sit in such fear. Trapped by not caring. Not having incentive. I need to spend time remembering why I am doing what I do everyday. And the why is hard when you don't care about....anything. The why can be so simple, short-term, long-term, for me, myself, I, friends, family, the world. Learn to relax the mind.


An optimized Analogy-Based Project Effort Estimation




Find appropriate # of closest projects (# of analogies)
Which adjustment technique


Conclusions: # of analogies was remarkably variable for each test project


Mohammad Azzeh, Yousef Elsheikh and Marwan Alseid, “An Optimized Analogy-Based Project Effort Estimation” International Journal of Advanced Computer Science and Applications(IJACSA), 5(4), 2014. http://dx.doi.org/10.14569/IJACSA.2014.050402 - See more at: http://thesai.org/Publications/ViewPaper?Volume=5&Issue=4&Code=IJACSA&SerialNo=2#sthash.EDi2Xwra.dpuf


Dataset Quality Assessment: An extension for analogy based effort estimation

ASSUMPTION: Similar projects in terms of attribute values will also be similar in terms of effort values.
Kendall row wise correlation : able to provide statistical evidence for attribute subset selection, statistical mechanism for identification of abnormal projects within a dataset, able to assess relevancy of nominal(categorical) attributes individually

APA

Azzeh, M. (2013). Dataset quality assessment: An extension for analogy based effort estimation. International Journal of Computer Science and Engineering Survey, 4, 1-21.

MLA

Azzeh, Mohammad. "Dataset Quality Assessment: An Extension for Analogy Based Effort Estimation." International Journal of Computer Science and Engineering Survey, 4 (2013): 1-21.





Comments

Post a Comment

Popular posts from this blog

some url/category/action_within_category create an instance of the category controller and then calls the action with that instance looks for template to display the action and result heading to /views/say and seeing if it finds a specific template for that action
Read more
a bunch of methods that can be used with any other class (MIX IN) a method is a group of several expressions(block) that can be referred by a name, passed arguments, associated with one or more objects higher order function takes a function as a param or returns as a result AREL relational algebra  returning the reply for the first comment, a table joining itself why is that easier in arel than sql, because sql qould require aliasing CONTEXT When    user logs out when    user logs in when    not valid or With TEST VALID EDGE NOT VALID ex  describe destroy do        context 1: WHEN resrouce found => 200        context 2: WHEN resource not found        context3 : WHEN resource not OWNED expect DONT SHOULD expect dont should expect dont should expect means in my mind i think i know what will come rather than should, if what comes isn't what i thought obl...
Read more

ADVERTISING STYLES

storytelling      lifestyle focus MARKETING OBJECTIVES AWARENESS Brand - those MORE likely to pay attention to ads and increase awareness Local - geographic proximity Reach - Max # ppl CONSIDERATION traffic - send more people to a destination ON or OFF facebook engagement - comments, shares, likes, event responses, offer claims app installs -  video views lead generation - collect lead info from ppl interested in your business CONVERSION conversions - valuable actions (ex. adding payment info, making purchase) product catalog sales - automatically show products from catalog based on target audience store visits - promote geographic business location to those nearby AUDIENCE SELECTION Geographic location age gender language Detailed targeting include/exclude :  DEMOGRAPHICS INTERESTS BEHAVIORS Connections Facebook pages - people who like, friends of people who like, exclude Apps - same as above Events - same as...
Read more