Hasil Pencarian  ::  Simpan CSV :: Kembali

"Multivariate Analysis for the Social Sciences provides clear guidelines combined with the insight needed to understand the methods and applications of multivariate statistics. This easy-to-follow book provides students in social, behavioral, and health science-whose focus is not primarily in mathematics-with an abundance of chapter-ending questions and answers, including: conceptual questions about the meaning of each method; questions that test the reader?s ability to carry out the computational procedures on simple datasets; and data analysis questions for using analytical packages to analyze both simplest case data and also to practice with more realistic datasets."

"Kejadian multivariat adalah kejadian-kejadian yang memiliki tidak hanya satu peristiwa yang mempengaruhi, tetapi bisa lebih banyak peristiwa yang memberi dampak pada peristiwa utamanya. Dampak yang dihasilkan dari suatu kejadian dapat berupa apa saja dan bisa diprediksi. Hal ini menyebabkan perlunya dibentuk sebuah model untuk memprediksi dampak dari sebuah kejadian sehingga dapat diambil keputusan penting berdasarkan kejadian tersebut. Saham merupakan salah satu contoh yang dapat direpresentasikan sebagai kejadian multivariat, seperti harga saham saat penutupan atau closing price, harga maksimal penutupan saham pada periode tertentu, dan durasi waktu (bulanan). Harga penutupan saham dan harga maksimal penutupan saham pada periode tertentu merupakan variabel acak kontinu yang masing-masing diasumsikan berdistribusi eksponensial dan truncated logistic. Durasi waktu (bulanan) merupakan variabel acak diskrit yang diasumsikan berdistribusi geometrik. Untuk mengakomodir kejadian multivariat yang melibatkan ketiga variabel acak tersebut digunakan distribusi trivariat yaitu, distribusi TETLG (Trivariate distribution with Exponential, Truncated Logistic, and Geometric marginals). Selanjutnya, untuk mengetahui pola hubungan antara ketiga variabel acak sebagai vektor respon dengan tiga kovariat yaitu, tingkat pengangguran, tingkat inflasi, dan tingkat obligasi 10 tahun, dikonstruksi sebuah Generalized Linear Model (GLM) untuk kejadian multivariat. Estimasi parameter model GLM kejadian multivariat, dilakukan menggunakan metode Maximum Likelihood. Sebagai implementasi pemodelan harga saham menggunakan GLM kejadian multivariat, diterapkan pada data harga penutupan saham dari Yahoo! Finance untuk periode 2 Januari 1958 hingga 17 April 2020. Berdasarkan uji likelihood ratio, diperoleh hasil bahwa hanya tingkat inflasi dan tingkat pengangguran yang memiliki pengaruh signifikan terhadap pemodelan harga saham.

---

Multivariate events are events that have not only one influencing event, but there can be more events that have an impact on the main event. The impact resulting from an event can be anything and can be predicted. This causes the need to establish a model to predict the impact of an event so that important decisions can be made based on the incident. Stocks are one example that can be represented as multivariate events, such as the closing price of a stock, the maximum closing price for a given period, the duration of time (monthly). The closing price of shares and the maximum closing price of shares in a certain period are continuous random variables which are assumed to have exponential distribution and truncated logistic respectively. Time duration (months) is a discrete random variable which is assumed to have a geometric distribution. To accommodate multivariate events involving the three random variables, a trivariate distribution is used, namely, the TETLG distribution (Trivariate distribution with Exponential, Truncated Logistic, and Geometric marginals). Furthermore, to determine the pattern of relationship between the three random variables as response vectors with three covariates, which are, the unemployment rates, the inflation rates, and the 10-year bond rates, a Generalized Linear Model (GLM) for multivariate events is constructed. Estimation of GLM model parameters for multivariate events was carried out using the Maximum Likelihood method. As an implementation of stock price modeling using GLM multivariate events, it is applied to closing stock price data from Yahoo! Finance for the period January 2, 1958 to April 17, 2020. Based on the likelihood ratio test, it was found that only the inflation rate and unemployment rate had a significant influence on stock price modelling."