Presenting a Hybrid Scheme of Machine Learning Combined with Metaheuristic Optimizers for Predicting Final Cost and Time of Project

Reza Bakhshi,Sina Fard Moradinia,Rasool Jani,Ramin Vafaei Poor 대한토목학회 2022 KSCE JOURNAL OF CIVIL ENGINEERING Vol.26 No.8

A primary task to a project manager is to ensure that the project proceeds timely and on the budget. Earned value management (EVM) is the most common method to evaluate and further predict a project in terms of time and cost. Because of poor accuracy in forecasting different durations, the application of the EVM for schedule performance prediction has been limited. Therefore, in this paper, the Kalman filter was used to analyze the data and compare the results with those of a few hybrid machine learning (HML) techniques. The innovation of this study comes in the face of the prediction of the project time and cost by efficient machine learning algorithms based on some directly measurable variables. For this purpose, 398 data points from five different dam projects were used to predict two output variables by hybrid schemes of multilayer perceptron (MLP) algorithm combined with the genetic algorithm (GA) and particle swarm optimization (PSO), herein referred to as MLP-GA and MLP-PSO, respectively. Four input variables, namely earned schedule (month), earned value ($), actual progress (%), and actual cost (%), were considered for time prediction, and the time (month) was considered as an input variable for cost prediction. The results showed that early warnings for the risk of delay in project schedule and cost were generated in months 1 and 8 by the MLP-PSO, months 1 and 11 by the MLP-GA, and rather months 7 and NV (no value) for the Kalman filter, indicating the fast operation coupled with high accuracy of the HML algorithms, as compared to the Kalman filter. In addition, the two HML algorithms were compared based on statistical error indices to determine the best one. After reviewing the results, the MLP-PSO was found to outperform the MLP-GA in terms of convergence rate and accuracy. To the best of our knowledge, none of the previous studies have used the proposed hybrid algorithms (MELM-PSO and MELM-GA) for forecasting, simultaneously, the time and cost as two critical parameters in the project management, indicating the novelty of our research.

Psychosocial Factors and Musculoskeletal Pain Among Rural Hand-woven Carpet Weavers in Iran

Reza Chaman,Roqayeh Aliyari,Farideh Sadeghian,Javad Vatani Shoaa,Mahmood Masoudi,Shiva Zahedi,Mohammad A. Bakhshi 한국산업안전보건공단 산업안전보건연구원 2015 Safety and health at work Vol.6 No.2

Background: Musculoskeletal pain (MSP) is a common and disabling problem among carpet weavers and is linked to physical and psychosocial factors of work. This study aimed to determine the prevalence of MSP, its psychosocial risk factors, and association of pain in each pair of anatomical sites among carpet weavers. Methods: A cross-sectional study was performed among 546 hand-woven carpet weavers in rural smallscale workshops of Iran. Data were collected by using parts of a standardized CUPID (Cultural and Psychosocial Influences on Disability) questionnaire focused on MSP in 10 body sites, including the lowback, neck, both right and left shoulders, elbows, wrists/hands, individual, physical and psychosocial risk factors. Statistical analysis was performed applying logistic regression models. Results: Prevalence of MSP in at least one body sitewas 51.7% over the past month. The most common sites were low back and right shoulder pain 27.4% and 20.1%, respectively. A significant difference was found between the mean number of painful anatomical sites and the level of education, age, physical loading at work, time pressure, lack of support, and job dissatisfaction. In pairwise comparisons, strongest association was found between pain in each bilateral anatomical site (odds ratio ¼ 11.6e35.3; p < 0.001). Conclusion: In home-based workshops of carpet weaving, psychosocial factors and physical loading were associated with MSP. This finding is consistent with studies conducted among other jobs. Considering the preventive programs, the same amount of attention should be paid to psychosocial risk factors and physical loading. Also, further longitudinal studies are needed to investigate the relationship of psychological factors.

Psychosocial Factors and Musculoskeletal Pain Among Rural Hand-woven Carpet Weavers in Iran

Chaman, Reza,Aliyari, Roqayeh,Sadeghian, Farideh,Shoaa, Javad Vatani,Masoudi, Mahmood,Zahedi, Shiva,Bakhshi, Mohammad A. Occupational Safety and Health Research Institute 2015 Safety and health at work Vol.6 No.2

Background: Musculoskeletal pain (MSP) is a common and disabling problem among carpet weavers and is linked to physical and psychosocial factors of work. This study aimed to determine the prevalence of MSP, its psychosocial risk factors, and association of pain in each pair of anatomical sites among carpet weavers. Methods: A cross-sectional study was performed among 546 hand-woven carpet weavers in rural small-scale workshops of Iran. Data were collected by using parts of a standardized CUPID (Cultural and Psychosocial Influences on Disability) questionnaire focused on MSP in 10 body sites, including the low-back, neck, both right and left shoulders, elbows, wrists/hands, individual, physical and psychosocial risk factors. Statistical analysis was performed applying logistic regression models. Results: Prevalence of MSP in at least one body sitewas 51.7% over the past month. The most common sites were low back and right shoulder pain 27.4% and 20.1%, respectively. A significant difference was found between the mean number of painful anatomical sites and the level of education, age, physical loading at work, time pressure, lack of support, and job dissatisfaction. In pairwise comparisons, strongest association was found between pain in each bilateral anatomical site (odds ratio = 11.6-35.3; p < 0.001). Conclusion: In home-based workshops of carpet weaving, psychosocial factors and physical loading were associated with MSP. This finding is consistent with studies conducted among other jobs. Considering the preventive programs, the same amount of attention should be paid to psychosocial risk factors and physical loading. Also, further longitudinal studies are needed to investigate the relationship of psychological factors.

Artificial Neural Network for Prediction of Distant Metastasis in Colorectal Cancer

Biglarian, Akbar,Bakhshi, Enayatollah,Gohari, Mahmood Reza,Khodabakhshi, Reza Asian Pacific Journal of Cancer Prevention 2012 Asian Pacific journal of cancer prevention Vol.13 No.3

Background and Objectives: Artificial neural networks (ANNs) are flexible and nonlinear models which can be used by clinical oncologists in medical research as decision making tools. This study aimed to predict distant metastasis (DM) of colorectal cancer (CRC) patients using an ANN model. Methods: The data of this study were gathered from 1219 registered CRC patients at the Research Center for Gastroenterology and Liver Disease of Shahid Beheshti University of Medical Sciences, Tehran, Iran (January 2002 and October 2007). For prediction of DM in CRC patients, neural network (NN) and logistic regression (LR) models were used. Then, the concordance index (C index) and the area under receiver operating characteristic curve (AUROC) were used for comparison of neural network and logistic regression models. Data analysis was performed with R 2.14.1 software. Results: The C indices of ANN and LR models for colon cancer data were calculated to be 0.812 and 0.779, respectively. Based on testing dataset, the AUROC for ANN and LR models were 0.82 and 0.77, respectively. This means that the accuracy of ANN prediction was better than for LR prediction. Conclusion: The ANN model is a suitable method for predicting DM and in that case is suggested as a good classifier that usefulness to treatment goals.

The Within-tree Variation in Wood Density and Mechanical Properties and Their Relationship in Juniperus polycarpos

Kiaei, Majid,Bakhshi, Reza,Saffari, Mohsen,Golkari, Sadegh Institute of Forest Science 2015 Journal of Forest Science Vol.31 No.4

The variations of wood density and mechanical properties of Juniperus polycarpos trees were studied in a natural forest in Iran. Sample disks were taken from each tree to examine wood density and mechanical properties (MOE and MOR) from pith to bark at breast height, 50%, and 75% of total tree height. The analysis of variance (ANOVA) indicated that radial position and height significantly affected all wood properties. The wood density, MOE and MOR were decreased along horizontal position from the pith to the bark and vertical direction from base upwards. Regression analysis showed that modulus of elasticity (MOE) and modulus of rupture (MOR) had a positive correlation with wood density.

Fruit Quality, Anthocyanin, and Cyanidin 3-Glucoside Concentrations of Several Blood Orange Cultivars Grown in Different Areas of Iran

Javad Fattahi,Reza Fotouhi,Davoud Bakhshi,Sirous Aghajanzadeh 한국원예학회 2009 Horticulture, Environment, and Biotechnology Vol.50 No.4

Fruit quality and its appearance are very important from consumer’s view point. The fruits of ‘Moro’, ‘Tarocco’, ‘Sanguinello’, and ‘Sanguine’ blood orange cultivars were harvested at the stage of commercial maturity in Kotra (North) and Darab (South) counties in Iran. After harvesting, the fruits were analyzed for Hunter value, total soluble solids (TSS) content, titratable acidity (TA), and concentrations of anthocyanin and cyanidin 3-glucoside. ‘Tarroco’ fruits from Darab region showed high Hunter L value (70.3). The oranges harvested in Kotra had a hue angel value of more than 80. Chroma values were at the highest amount in ‘Tarocco’ with 70.6 values. Generally, TSS was higher in the cultivars grown in south than north. ‘Sanguinella’ and ‘Sanguine’ fruits had the highest TA in both regions. Cyanidin 3-glucoside contents of ‘Tarocco’ (South) and ‘Moro’ (North) were 368.2 and 196.05 ㎎ㆍL?¹, respectively, which were higher than the others. However, ‘Sanguinello’ and ‘Sanguine’ fruits grown in the southern region contained higher amount of cyanidin 3-glucoside than those in the northern area. In this experiment, we found that climate was the major influence on main component and anthocyanin of fruits. ‘Tarocco’ and ‘Moro’ fruits had high quality in south and north counties, respectively.

The Within-tree Variation in Wood Density and Mechanical Properties and Their Relationship in Juniperus polycarpos

Majid Kiaei,Reza Bakhshi,Mohsen Saffari,Sadegh Golkari 강원대학교 산림과학연구소 2015 Journal of Forest Science Vol.31 No.4

The variations of wood density and mechanical properties of Juniperus polycarpos trees were studied in a natural forest in Iran. Sample disks were taken from each tree to examine wood density and mechanical properties (MOE and MOR) from pith to bark at breast height, 50%, and 75% of total tree height. The analysis of variance (ANOVA) indicated that radial position and height significantly affected all wood properties. The wood density, MOE and MOR were decreased along horizontal position from the pith to the bark and vertical direction from base upwards. Regression analysis showed that modulus of elasticity (MOE) and modulus of rupture (MOR) had a positive correlation with wood density.

Ascorbic Acid, Anthocyanins, and Phenolics Contents and Antioxidant Activity of Ber, Azarole, Raspberry, and Cornelian Cherry Fruit Genotypes Growing in Iran

Abouzar Hashempour,Reza Fotuhi Ghazvini,Davood Bakhshi,Mahmood Ghasemnezhad,Mohammad Sharafti,Hamid Ahmadian 한국원예학회 2010 Horticulture, Environment, and Biotechnology Vol.51 No.2

The attention paid by consumers to the lesser known fruits has recently been increased. These fruits have unusual flavors and qualities, and many of which are rich in antioxidant substances such as phenolics specifically anthocyanins. In the present study, biochemical and antioxidant properties of ber (Ziziphus spina-christi), azarole (Crataegus azarolus), raspberry (Rubus hyrcanus), and cornelian cherry (Cormus mas) fruits with varied pigmentation obtained from Iranian native populations were studied. Total phenolics and ascorbic acid contents varied among these populations. Raspberry had the highest total phenolics and ascorbic acid content with 454.7 and 143.2 ㎎?100 g<SUP>-1</SUP> fresh weight, respectively. Consequently, it showed the highest total anthocyanin (182.6 ㎎ pelargonidin-3-glucoside equivalent 100 g<SUP>-1</SUP> fresh weight) content and antioxidant activity (82.3% scavenged 1,1-diphenyl-2-picrylhydrazyl radical). While the levels of total soluble solids and total acidity was highest in cornelian cherry population (12.3 and 9.9%, respectively). Antioxidant capacity was highly correlated with total anthocyanin and total phenolics content (r² = 0.78 and 0.72, respectively), whereas linear correlation between total antioxidant capacity with ascorbic acid was lower (r² = 0.43). The present study showed that the native raspberry fruit is an extremely rich source of antioxidants in comparison with ber, azarole, and cornelian cherry.

The effect of carbon fiber length on the microstructure, selected mechanical, wear, and thermal conductivity of Cf/SiC composite fabricated via spark plasma sintering (SPS) method

Ghasemi Arman,Ramazani Mazaher,Bakhshi Saeed Reza,Al-Khafaji Ali Hussein Demin,Zahabi Saeed,Loghman Estarki Mohammad Reza,Zamani Abbas 한국세라믹학회 2023 한국세라믹학회지 Vol.60 No.4

This research aims to fabricate and develop a composite brake disc made of carbon/silicon carbide. For this purpose, the first silicon carbide nanoparticles were ultrasonicated with alumina sintering aid and carbon fi ber (between 10 and 20%) with different lengths of 3 mm, 10 mm, and 15 mm. Next, the fi nal Cf/SiC composite bulk was made-up via the spark plasma sintering (SPS) method. After that, the effect of different lengths of carbon fiber was explored on the porosity, coefficient of friction, fracture toughness, thermal conductivity, and microhardness of samples. The results showed that the density of sintered samples with a carbon fiber length of 10 mm was higher than that of samples with fiber lengths of 3 and 15 mm. Also, the hardness (25.79 GPa) and fracture toughness (5.72 MPa.m 1/2) of this sample were higher than those of the samples sintered with carbon fiber lengths of 3 and 15 mm. Further, this sample with the maximum density showed a coefficient of friction (COF) of 0.43. Since for aeroplane brake discs, the COF should be between 0.3 and 0.4, and the porosity of sam- ples should range within 3–5%, the samples sintered with the carbon fiber length of 10 mm had the nearby features to the preferred air brake disc indices.

Preparation and Improved Friction and Wear Performance of the Nano-MoSi2 Coating on Ni Substrate by Plasma Spraying

Mohammad Erfanmanesh,Saeed Reza Bakhshi,Mohammad Reza Pakmanesh,Masoud Barekat 대한금속·재료학회 2021 METALS AND MATERIALS International Vol.27 No.6

In this work, the wear performance of nano- and ordinary MoSi2plasma-sprayed coating on Ni substrate are compared. MoSi2nanopowder is prepared by ball milling followed by heat treatment processes. Atmospheric plasma spraying withargon atmosphere protection has been used to deposit the powders onto the nickel substrate. Ni substrate, MoSi2coating,and nano-MoSi2 coating were subjected to reciprocating wear test. Wear tests were done on a pin on plate machine by slidingMoSi2coating and substrates without coating against 52100 steel as the pin, at a temperature of 27 °C in dry condition. A friction coefficient of 0.18 ± 0.1 is obtained for nanostructured MoSi2coating, whereas a friction coefficient of 0.2 ± 0.1is obtained for conventional MoSi2coating under normal load of 20 N. Wear performance of nanostructured MoSi2coatingwas better as compared to the conventional MoSi2coating. The three observed dominant types of wear mechanisms wereadhesive, oxidation and abrasive. Morphologies and phases of coating, wear tracks, wear debris and counterpart steel pinswere evaluated by SEM with EDS and X-ray diffraction. The microhardness and adhesion strength of nanostructured MoSi2coating are determined by 1000 HV and 27 MPa, respectively.