Electronic Theses and Dissertation

Globalisasi telah mendorong perekonomian berbagai negara di dunia, salah satunya adalah negara di benua asia yang menunjukkan pertumbuhan ekonomi yang pesat dalam beberapa dekade terakhir. nilai tukar mata uang menjadi indikator penting dalam perkembangan sistem perekonomian antara negara ke arah yang lebih terbuka, dengan dolar amerika serikat sebagai salah satu mata uang internasional. pergerakan nilai tukar mata uang mampu mempengaruhi stabilitas perekonomian suatu negara. namun, perubahan nilai tukar antar negara sulit dilihat pergerakannya secara menyeluruh, menciptakan ketidakpastian dalam perdagangan internasional, seperti terlihat dari melemahnya sebagian besar mata uang asia terhadap dolar as (usd) ditengah penguatan ekonomi as. untuk memahami dinamika ini, penerapan k-means clustering pada data time series nilai tukar mata uang di asia dapat menghasilkan pola pengelompokan berdasarkan kemiripan perilaku pergerakan, sehingga dapat dilihat pola pergerakan nilai tukar mata uang di asia terhadap dolar as (usd) secara regional. pengelompokan mata uang berdasarkan nilai tukarnya terhadap dolar as (usd) dilakukan dengan menggunakan k-means clustering menggunakan tiga variasi pengukuran jarak yaitu dynamic time warping (dtw), soft-dynamic time warping (soft-dtw) dan euclidean. sedangkan penentuan jumlah cluster yang optimal diuji menggunakan silhouette index. hasil penelitian menunjukkan pengelompokan nilai tukar mata uang terhadap dolar as (usd) paling optimal adalah menggunakan jarak soft-dtw dengan jumlah cluster 4. pemilihan jarak soft-dtw dengan jumlah cluster 4 sebagai hasil cluster optimal berdasarkan nilai uji validitas silhouette index yang menghasilkan nilai paling tinggi dan mendekati nilai +1 yaitu sebesar 0,618. berdasarkan 4 cluster hasil pengelompokan, pada cluster 1 terdapat 20 mata uang. cluster 2 terdapat 4 mata uang. cluster 3 terdapat 14 mata uang, dan cluster 4 terdapat 9 mata uang.

Globalization has driven the economies of various countries in the world, one of which is a country on the Asian continent that have shown rapid economic growth in the last few decades. Currency exchange rates are an important indicator in the development of economic systems between countries towards a more open direction, with the United States Dollar as one of the international currencies. The movement of currency exchange rates can affect the stability of a country's economy. However, changes in exchange rates between countries are difficult to see the overall movement, creating uncertainty in international trade, as seen from the weakening of most Asian currencies against the US dollar (USD) amid the strengthening of the US economy. To understand these dynamics, the application of k-means clustering on time series data of currency exchange rates in Asia can produce clustering patterns based on similar movement behavior, so that regional patterns of currency exchange rate movements in Asia against the US dollar (USD) can be seen. Clustering of currencies based on their exchange rates against the US dollar (USD) is done using k-means clustering using three variations of distance measurements, namely Dynamic Time Warping (DTW), Soft-Dynamic Time Warping (Soft-DTW) and Euclidean. While the determination of the optimal number of clusters is tested using the silhouette index. The results showed that the most optimal clustering of currency exchange rates against the US dollar (USD) is using the Soft-DTW distance with the number of clusters 4. Selection of the Soft-DTW distance with the number of clusters 4 as the optimal cluster result based on the validity test value of the silhouette index which produces the highest value and is close to the value of +1 which is 0.618. Based on 4 clusters of clustering results, in cluster 1 there are 20 currencies. Cluster 2 there are 4 currencies. Cluster 3 there are 14 currencies, and cluster 4 there are 9 currencies.