Gülseren Karabay, Muhammet Burak Kılıç, Kazim Saricoban, G. K. Günaydin
{"title":"利用人工神经网络自回归模型预测土耳其服装出口","authors":"Gülseren Karabay, Muhammet Burak Kılıç, Kazim Saricoban, G. K. Günaydin","doi":"10.35530/it.074.02.202265","DOIUrl":null,"url":null,"abstract":"Foreign trade is significant for open economies and has a critical place in the development of national economies in a\nglobally competitive environment. Export has a key role as an important component in foreign trade transactions. In this\nstudy, Turkey's exports of HS-61 “Apparel and clothing accessories knitted or crocheted” and HS-62 “Apparel and\nclothing accessories not knitted or crocheted” products were examined. Turkey's exports of these HS codes to seven\ncountries, which are mostly exported, EU27, OECD and the world were estimated for 2020–2025 using artificial neural\nnetworks (ANNs). The model was validated by measuring forecast errors (RMSE, MAE, and MAPE) to ensure that the\nmodel can recreate satisfactory results. The results reveal that the ANNs predicted the exports to the selected countries\naccurately. According to the findings from the NNAR (1,1) model, Turkey's exports of HS-61 coded products to Italy, the\nUK, France and EU-27 are expected to increase year-on-year from 2020 to 2025, while exports to the USA, Netherlands\nand Spain are expected to decrease. Exports from Turkey to Germany, the world and the OECD are projected to\ndecrease first and then increase. In the NNAR (2,2) model, Turkey's exports to Italy, the UK and France are expected\nto increase year on year, while exports to other countries are generally estimated to decrease first and then increase.\nIn the estimation made for the HS-62 product group with NNAR (1,1) model, it is predicted that TURKEY’s exports to\nItaly, the USA, the UK, France and Spain will increase year by year from 2020 to 2025. It is estimated that TURKEY's\nexports to the Netherlands, Germany, World, OECD and EU-27 will decrease year by year in the same period. In the\nNNAR (2,2) model, it is predicted that TURKEY's exports to UK, France and Spain will increase year by year, while\nexports to EU-27, OECD and World will decrease. However, TURKEY's HS-62 exports to Italy, the USA, the Netherlands\nand Germany are expected to follow a fluctuating course.","PeriodicalId":13638,"journal":{"name":"Industria Textila","volume":" ","pages":""},"PeriodicalIF":1.0000,"publicationDate":"2023-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Forecasting of Turkey's apparel exports using artificial neural network\\nautoregressive models\",\"authors\":\"Gülseren Karabay, Muhammet Burak Kılıç, Kazim Saricoban, G. K. Günaydin\",\"doi\":\"10.35530/it.074.02.202265\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Foreign trade is significant for open economies and has a critical place in the development of national economies in a\\nglobally competitive environment. Export has a key role as an important component in foreign trade transactions. In this\\nstudy, Turkey's exports of HS-61 “Apparel and clothing accessories knitted or crocheted” and HS-62 “Apparel and\\nclothing accessories not knitted or crocheted” products were examined. Turkey's exports of these HS codes to seven\\ncountries, which are mostly exported, EU27, OECD and the world were estimated for 2020–2025 using artificial neural\\nnetworks (ANNs). The model was validated by measuring forecast errors (RMSE, MAE, and MAPE) to ensure that the\\nmodel can recreate satisfactory results. The results reveal that the ANNs predicted the exports to the selected countries\\naccurately. According to the findings from the NNAR (1,1) model, Turkey's exports of HS-61 coded products to Italy, the\\nUK, France and EU-27 are expected to increase year-on-year from 2020 to 2025, while exports to the USA, Netherlands\\nand Spain are expected to decrease. Exports from Turkey to Germany, the world and the OECD are projected to\\ndecrease first and then increase. In the NNAR (2,2) model, Turkey's exports to Italy, the UK and France are expected\\nto increase year on year, while exports to other countries are generally estimated to decrease first and then increase.\\nIn the estimation made for the HS-62 product group with NNAR (1,1) model, it is predicted that TURKEY’s exports to\\nItaly, the USA, the UK, France and Spain will increase year by year from 2020 to 2025. It is estimated that TURKEY's\\nexports to the Netherlands, Germany, World, OECD and EU-27 will decrease year by year in the same period. In the\\nNNAR (2,2) model, it is predicted that TURKEY's exports to UK, France and Spain will increase year by year, while\\nexports to EU-27, OECD and World will decrease. However, TURKEY's HS-62 exports to Italy, the USA, the Netherlands\\nand Germany are expected to follow a fluctuating course.\",\"PeriodicalId\":13638,\"journal\":{\"name\":\"Industria Textila\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":1.0000,\"publicationDate\":\"2023-05-02\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Industria Textila\",\"FirstCategoryId\":\"88\",\"ListUrlMain\":\"https://doi.org/10.35530/it.074.02.202265\",\"RegionNum\":4,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"MATERIALS SCIENCE, TEXTILES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Industria Textila","FirstCategoryId":"88","ListUrlMain":"https://doi.org/10.35530/it.074.02.202265","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"MATERIALS SCIENCE, TEXTILES","Score":null,"Total":0}
Forecasting of Turkey's apparel exports using artificial neural network
autoregressive models
Foreign trade is significant for open economies and has a critical place in the development of national economies in a
globally competitive environment. Export has a key role as an important component in foreign trade transactions. In this
study, Turkey's exports of HS-61 “Apparel and clothing accessories knitted or crocheted” and HS-62 “Apparel and
clothing accessories not knitted or crocheted” products were examined. Turkey's exports of these HS codes to seven
countries, which are mostly exported, EU27, OECD and the world were estimated for 2020–2025 using artificial neural
networks (ANNs). The model was validated by measuring forecast errors (RMSE, MAE, and MAPE) to ensure that the
model can recreate satisfactory results. The results reveal that the ANNs predicted the exports to the selected countries
accurately. According to the findings from the NNAR (1,1) model, Turkey's exports of HS-61 coded products to Italy, the
UK, France and EU-27 are expected to increase year-on-year from 2020 to 2025, while exports to the USA, Netherlands
and Spain are expected to decrease. Exports from Turkey to Germany, the world and the OECD are projected to
decrease first and then increase. In the NNAR (2,2) model, Turkey's exports to Italy, the UK and France are expected
to increase year on year, while exports to other countries are generally estimated to decrease first and then increase.
In the estimation made for the HS-62 product group with NNAR (1,1) model, it is predicted that TURKEY’s exports to
Italy, the USA, the UK, France and Spain will increase year by year from 2020 to 2025. It is estimated that TURKEY's
exports to the Netherlands, Germany, World, OECD and EU-27 will decrease year by year in the same period. In the
NNAR (2,2) model, it is predicted that TURKEY's exports to UK, France and Spain will increase year by year, while
exports to EU-27, OECD and World will decrease. However, TURKEY's HS-62 exports to Italy, the USA, the Netherlands
and Germany are expected to follow a fluctuating course.
期刊介绍:
Industria Textila journal is addressed to university and research specialists, to companies active in the textiles and clothing sector and to the related sectors users of textile products with a technical purpose.