Napredna pretraga

Pregled bibliografske jedinice broj: 472659

Predicting cost of prefabricated housing using neural networks


Vukomanović, Mladen; Kararic, Mirsad; Radujković, Mladen
Predicting cost of prefabricated housing using neural networks // Proceedings of the PM-05> Fifth Scientific Conference on Project Management / John-Paris Pantouvakis (ur.).
Atena: Center for Construction Innovation, National Technical University of Athens, 2010. (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
Predicting cost of prefabricated housing using neural networks

Autori
Vukomanović, Mladen ; Kararic, Mirsad ; Radujković, Mladen

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the PM-05> Fifth Scientific Conference on Project Management / John-Paris Pantouvakis - Atena : Center for Construction Innovation, National Technical University of Athens, 2010

ISBN
978-960-254-690-1

Skup
PM-05> Fifth Scientific Conference on Project Management

Mjesto i datum
Heraklion, Grčka, 29-31.05.2010

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Prefabricated housing; prediction; cost; neural networks; model; construction
(Prefabricated housing; prediction; cost; neural networks; model; construction+)

Sažetak
Political and economic pressures have become an aggravating circumstance for construction companies in achieving basic project management criteria, i.e. cost, time and scope. The construction's low performance only stresses out the need for improving current practices - especially in regard to cost. Therefore, we sought to find a critical set of variables for predicting total cost of prefabricated housing. We applied neural networks on data from more than 30 projects and thus have identify 17 critical variables for the cost prediction. The model was verified on 28 buildings with following performances: 85.7% of predicted values had the deviation lower 5%, while 10.7% had the deviation lower than 10%, in relation to the actual cost. After validating the model on data from 3 buildings, that had been new to the network, the performances were as follows: 83.8% of predicted values had the deviation lower 5%, while 12.9% had the deviation lower than 10%. Therefore this model showed to be very robust. Furthermore, this study has also demonstrated a more efficient and effective way of predicting total cost of building. Thus construction companies can influence project performance during project early phases, and acquire more competitive position on the market. Conclusion brings guidelines for use of the model and gives recommendation for its further development.

Izvorni jezik
Engleski

Znanstvena područja
Građevinarstvo



POVEZANOST RADA


Projekt / tema
082-0822156-2993 - Upravljanje rizikom i promjenama u projektno usmjerenom građevinskom poslovanju (Mladen Radujković, )
082-0822156-2998 - Upravljanje ljudskim potencijalima u građevinarstvu (Anita Cerić, )

Ustanove
Građevinski fakultet, Zagreb