• Geomechanics and Engineering
• /
• v.25 no.1
• /
• pp.1-16
• /
• 2021

Machine learning models have been widely used for landslide susceptibility assessment (LSA) in recent years. The large number of inputs or conditioning factors for these models, however, can reduce the computation efficiency and increase the difficulty in collecting data. Feature selection is a good tool to address this problem by selecting the most important features among all factors to reduce the size of the input variables. However, two important questions need to be solved: (1) how do feature selection methods affect the performance of machine learning models? and (2) which feature selection method is the most suitable for a given machine learning model? This paper aims to address these two questions by comparing the predictive performance of 13 feature selection-based machine learning (FS-ML) models and 5 ordinary machine learning models on LSA. First, five commonly used machine learning models (i.e., logistic regression, support vector machine, artificial neural network, Gaussian process and random forest) and six typical feature selection methods in the literature are adopted to constitute the proposed models. Then, fifteen conditioning factors are chosen as input variables and 1,017 landslides are used as recorded data. Next, feature selection methods are used to obtain the importance of the conditioning factors to create feature subsets, based on which 13 FS-ML models are constructed. For each of the machine learning models, a best optimized FS-ML model is selected according to the area under curve value. Finally, five optimal FS-ML models are obtained and applied to the LSA of the studied area. The predictive abilities of the FS-ML models on LSA are verified and compared through the receive operating characteristic curve and statistical indicators such as sensitivity, specificity and accuracy. The results showed that different feature selection methods have different effects on the performance of LSA machine learning models. FS-ML models generally outperform the ordinary machine learning models. The best FS-ML model is the recursive feature elimination (RFE) optimized RF, and RFE is an optimal method for feature selection.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.10 no.4
• /
• pp.64-69
• /
• 2011

Magnesium alloy has been known as lightweight material in automobile and electronic industry with aluminum alloy, titanium alloy and plastic material. Friction welding is useful to join different kinds of metals and nonferrous metals they are difficult to be joined by such as gas welding, resistance welding and electronic beam welding. In this study, friction welding was performed to investigate optimization process of Mg alloy with a 20mm diameter solid bar. For that, the orthogonal array $(L_{9}(3^{4}))$ was used that contained four factors and each factor had three levels. Control factors were heating pressure, heating time, upsetting pressure and upsetting time. Also tensile tests were carried out to measure mechanical properties for welded conditions. The levels of heating pressure and upsetting pressure used were 15, 25, 35MPa, and 30, 50, 70MPa, respectively. In addition those of heating time and upsetting time were 0.5, 1, 1.5 sec and 3, 4, 5 sec., respectively, rotating speed of 2000rpm. From the experimental results, optimization condition was estimated as follows; heating pressure=35MPa, upsetting pressure=70MPa, heating time=1.5sec, upsetting time=3sec.

• Journal of Korea Foundry Society
• /
• v.17 no.2
• /
• pp.145-150
• /
• 1997

The striations parallel to the casting direction are the characteristic defects occurring during the unidirectional solidification of nonferrous metals such as pure Al and Sn. It has been known that the defects affect the recrystallization temperature as well as the mechanical and electrical properties of Al. However, it has not been still known about the effect of casting factors such as the casting speed, rod diameter of cast metals, and alloying elements on the occurrence of striations. In the present study, these casting factors affecting the occurrence of striation were investigated by adopting the quantitative examination of striations. Also, the relation between striation defects and feathery grain, which are very similar to the striation defects, was studied.

• Transactions of Materials Processing
• /
• v.19 no.5
• /
• pp.277-282
• /
• 2010

Magnesium alloy has been known as lightweight material in automobile and electronic industry with aluminum alloy, titanium alloy and plastic material. Friction welding is useful to join various metals and nonferrous metals that are difficult to join by such as gas welding, resistance welding and electronic beam welding. In this study, friction joining was performed to investigate mechanical properties of Mg alloy with 20mm diameter solid bar. Also the optimal joining conditions for its application were determined on the basis of tensile test, and hardness survey. The joining parameters were chosen as heating pressure, heating time, upsetting pressure, and upsetting time. Heating and upsetting pressure were executed under the range of 10~40MPa and 20~80MPa, respectively. From the experimental results, optimal joining conditions were determined as follows; rotating speed=2000rpm, heating pressure=35MPa, upsetting pressure=70MPa, heating time=1sec, upsetting time=5sec. Also the hardness of jointed boundary showed as HV50 which was similar to that of base metal at the optimal condition, and it was supposed that zone of HAZ was 8mm. Finally two materials were strongly mixed at interface part to show a well-combined microstructure without particle growth or any defect.

• Journal of the Korean Chemical Society
• /
• v.29 no.2
• /
• pp.178-182
• /
• 1985

The purpose of this study was to develop a separation method of Zn(II), Cu(II) and Mg(II), by ion exchange chromatography using cation exchange resion (Dowex 50w${\times}$8, 80-100 mesh) and anion exchange (Amberlite IRA-400). Ion exchange resions were packed into 25 ${\times}$ 2cm ID column and flow rate was controlled to 0.30 ml/min. Good eluents for separation of nonferrous metal ions such as Zn(II), Cu(II), Mg(II) were as follow: 0.5M $NaNO_3$ (pH 3.1), 0.2~0.5M HCl + 50~60% Acetone, and 1M HAc + 0.1M NaAcf(pH 3.7) aqueous solution. The mixed solution of 0.1M NaAc(pH 3.7), 0.5M HCl + 50% Acetone were found to be the best eluent for step elution. Analysis of metals were determined by atomic absorption spectrophotometer. In addition, separated Zn(II) fraction was obtained by eluted with 0.12N HCl and 1.5N $NH_4OH$ aqueous solution. This solution was titrated by the E. D. T. A.

• Journal of Korean Society for Atmospheric Environment
• /
• v.20 no.5
• /
• pp.685-695
• /
• 2004

Size-segregated measurements of aerosol composition are used to estimate the transport of natural and anthropogenic aerosols at Gosan site during May 2002. The results of measurement show that not only soil dust but also anthropogenic aerosols, including sulfur and enriched trace metals such as Pb, Zn, Cu, are transported to Gosan. This study combines the size- and time-resolved aerosol composition measurements with factor analysis in order to identify some source materials. As a result, coarse particles (2.5${\mu}m$~12${\mu}m$) are influenced by soil, sea-salt, coal, coal combustion, and nonferrous sources. But fine particles have different sources. The fine particles, which the diameter is from 0.56${\mu}m$ to 2.5${\mu}m$, are more affected by road dust, oil combustion, industry. municipal incineration, and ferrous metal sources. The very fine particles, from 0.09${\mu}m$ to 0.56${\mu}m$, mainly supplied by biomass burning, oil combustion, nonferrous and ferrous metal sources.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.14 no.6
• /
• pp.12-17
• /
• 2000

IN induction heating system the high frequency operation allows a rapid response to current fluctuation in the inverter and result in improved welding quality. To work induction heating of nonferrous metals, a welding power supply is need high working frequency and high power. This paper is shown design technique for increasing working frequency in induction heating for welding coppers. A series-parallel resonate inverter consists of H-type bridges, each of whose arms is composed of a combination of two parallel IGBTs. Inverter operating with the fixed frequency is controlled by pulse width modulation (PWM). As switching adapted the Zero-Voltage Switching technique to reduce switching losses the system is high efficiency. The propose inverter has feature which is high efficiency for very wide load variations with a narrow range of duty cycle ratio control and load short circuit capability. Detailed experimental results obtained from a 48[V] output, 500[W] experimental inverter are presented to verify the concept.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 1997.10a
• /
• pp.284-291
• /
• 1997

Many papers have been conducted on the applications of honeycomb sandwich maintenance, and to improve the high speed and light weight in rolling stocks, aircrafts and so on. The end door of rolling stock is generally made of rolled steel or stainless steel. Thus, the weight of these materials are heavier than of nonferrous metals and thermal deformation by welding or complexity of manufacturing process is occurred. Therefore, this paper is aimed to develop the light weight by application of end door which is made of aluminum honeycomb sandwich panels in rolling stocks and to propose the standards of design and evaluation for its adhesively bonded strength.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.11 no.3
• /
• pp.41-47
• /
• 2012

Surface roughness is widely used as an index for processing degree of accuracy. Recently, regression analysis to predict the machining results are actively used to characterize a cutting operations. In the past, diamond machining had been used for ultra precision cutting operation, but now industrial diamond tools like PCD(Polycrystaline Diamond) has been widely used in ultraprecision machining of nonferrous metals. In this study, the authors focus on the effect of PCD tool property on the surface roughness of different types of aluminum alloy after cutting process by CNC operated lathe. Based on the regression analysis approach on a surface roughness data obtained by experiment, predictive analysis of surface roughness is effective to achieve better surface quality.

• Proceedings of the IEEK Conference
• /
• 2001.10a
• /
• pp.770-775
• /
• 2001

Effects of magnetic field and carrier gas velocity on the magnetic separation of FCC catalyst by a high gradient magnetic separator were studied. The activities of the equilibrium catalyst, the magnetic particles and the nonmagnetic particles were evaluated in a fixed bed microreactor The results showed that heavy metal contaminated catalyst can be selectively separated by means of high gradient magnetic separation at magnetic field 0.5T and carrier gas velocity 0.3m.s$^{-1}$ , and lightly metal contaminated catalyst retained high catalytic activity.