Several litigation outcome prediction approaches are reviewed in the construction disputes area. ... more Several litigation outcome prediction approaches are reviewed in the construction disputes area. The reviewed approaches include artificial neural networks, boosted decision trees, particle swarm optimization, split-step particle swarm optimization, case based reasoning and integrated prediction model. The integrated prediction model outweighs the rest of the approaches, achieving a prediction rate of 91% using 132 training litigation cases only. Although there are over 45 attributes that might affect a construction litigation case, it is observed that 10 to 15 attributes would be sufficient to predict the outcome of litigation in the area of construction disputes. It is found that the most important attributes are type of contract, type of parties involved in the dispute, directed employer changes and liquidated damages.
Several litigation outcome prediction approaches are reviewed in the construction disputes area. ... more Several litigation outcome prediction approaches are reviewed in the construction disputes area. The reviewed approaches include artificial neural networks, boosted decision trees, particle swarm optimization, split-step particle swarm optimization, case based reasoning and integrated prediction model. The integrated prediction model outweighs the rest of the approaches, achieving a prediction rate of 91% using 132 training litigation cases only. Although there are over 45 attributes that might affect a construction litigation case, it is observed that 10 to 15 attributes would be sufficient to predict the outcome of litigation in the area of construction disputes. It is found that the most important attributes are type of contract, type of parties involved in the dispute, directed employer changes and liquidated damages.
Uploads