dự báo mô hình arima

... chọn là Mô hình 1. 2.4. Dự báo bằng mô hình ARIMA Mô hình ARIMA (0,1,1)(0,1,1) 12 tt uLLYLL )1)(1()1)(1( 12 11 12 Θ−−=−− θ (2.8) Tuy nhiên, để sử dụng một mô hình được nhận dạng cho dự báo, cần ... và dự báo lạm phát. Mục đích của bài viết này nhằm ứng dụng mô hình ARIMA với phương pháp Box-Jenkins để dự báo lạm phát ở Việt Nam. George Box và Gwilym Jenkins (1976) đã nghiên cứu mô hình ... của dữ liệu gốc (Hình 2.1) và đồ thị tương quan của dữ liệu sau khi biến đổi sai phân (Hình 2.2). Các mô hình được nhận dạng như sau:  Mô hình ARIMA( 0,1,1)(0,1,1) 12 (Mô hình 1) hoặc (1...

Ngày tải lên: 30/10/2012, 14:19

... sách sẽ làm cho sai số dự báo tăng cao hơn. Do đó kết quả của mô hình vẫn chỉ mang tính chất tham khảo nhiều hơn. Tuy nhiên có thể nói mô hình ARIMA là một mô hình tốt để dự báo trong ngắn hạn. ... bảng 1 ta thấy mô hình ARIMA( 0,1,1) là mô hình phù hợp với R nhất. Mô hình số quan sát 2 (4) AIC Arima( 0,1,1) 476 0.129571 -4.976207 Arima( 1,1,1) 476 0.104665 -4.966832 Arima( 1,1,0) 476 ... sử dụng mô hình ARIMA và phương pháp Box-jenkins để dự báo chỉ số VnIndex trong ngắn hạn căn cứ vào chuỗi dữ liệu quá khứ. George Box và Gwilym Jenkins (1976) đã nghiên cứu mô hình ARIMA (Autoregressive...

Ngày tải lên: 13/04/2013, 13:06

... mô hình và nhận thấy mô hình ARIMA( 2,1,3) dường như là phù hợp nhất. Bước 4: Dự báo Một trong số các lý do về tính phổ biến của phương pháp lập mô hình ARIMA là thành công của nó trong dự báo. ... thể nói phần dư của mô hình là nhiễu trắng và mô hình phù hợp. Nếu mô hình không phù hợp ta quay lại bước 1. Tuy nhiên, một chuỗi dữ liệu có thể phù hợp với nhiều mô hình ARIMA khác nhau, do ... tra cụ thể. Để dự báo trong Eviews trước tiên ta mở rộng dữ liệu: Procs/ Change Worfile range. Tại cửa số Forecast của hàm cần dự báo, ta chọn khoảng dự báo, trong ví dụ này ta dự báo từ giá trị...

Ngày tải lên: 25/07/2013, 10:57

Ngày tải lên: 07/08/2013, 19:25

... mô hình ARIMA để dự báo o Nếumẫudự liệu thay đổicầnphải ướclượng lại mô hình hoặcxâydựng mộtmôhìnhmới 10 Phùng Thanh Bình CHIẾN LƯỢC XÂY DỰNG MÔ HÌNH ARIMA z Bước1: Xácđịnh mô hình Giả sử mô ... Box-Jenkins 3. Mô hình tự hồi quy 4. Mô hình bình quân di động 5. Mô hình bình quân di động tự hồiquy 6. Chiếnlượcxâydựng mô hình ARIMA MÔ HÌNH ARIMA 11 Phùng Thanh Bình CHIẾN LƯỢC XÂY DỰNG MÔ HÌNH ARIMA z ... phảixácđịnh mô hình mới ∑ = − += m 1k 2 k m kn (e)r 2)n(n Q Phùng Thanh Bình CHIẾN LƯỢC XÂY DỰNG MÔ HÌNH ARIMA z Bước 4: Dự báo o Sau khi có một mô hình phù hợpcóthể thựchiện dự báo cho mộthoặcmộtsố...

Ngày tải lên: 14/03/2014, 21:23

... U t Bước 4: Dự báo Những dự báo ngắn hạn về lượng khách du lịch quốc tế đến Việt Nam dựa trên mô hình ARIMA (12, 1,12) được trình bày trong Bảng 5. Bảng 5 cho thấy số liệu dự báo lượng khách ... hình tốt (Wang & Lim, 2005). Bước 4: Dự báo: Dựa trên phương trình của mô hình ARIMA, tiến hành xác định giá trị dự báo điểm và khoảng tin cậy của dự báo. Bảng 1. Lượng khách quốc tế đến ... Box-Jenkins để xây dựng mô hình ARIMA cho dự báo lượng khách quốc tế đến Việt Nam dựa trên số liệu công bố hàng tháng của Tổng cục Du lịch Việt Nam. Kết quả cho thấy trong số các mô hình ước lượng...

Ngày tải lên: 24/03/2014, 23:21

... m Z t =Y t -Y t-m 2. Nhận dạng mô hình Mô hình ARIMA (hay còn gọi là phương pháp Box-Jenkin) Nhận dạng mô hình tức là xác định p, d, q trong ARIMA( p,d,q) p: dựa vào SPAC q: dựa vào SAC d: dựa vào số lần lấy ... Ví dụ dự báo giá gạo 1. Dữ liệu Hình 1 2. Xem chuỗi Rice có dừng không? 2 Hình 14 Hình 15 10 Hình 17 12 Hình 13 Như vậy, sai số của mô hình ARIMA( 1,1,1) là một chuỗi dừng ... của chuỗi này. Thử xem đồ thị Correlogram của chuỗi sai phân bậc 1 Hình 7 5 Hình 18 13 Hình 9 Hình 10 7 Dự báo bằng mô hình ARIMA (AutoRegressive Integrated Moving Average) 1. Tính dừng và...

Ngày tải lên: 25/03/2014, 09:30

... ta có mô hình ARIMA( p,1,q) Với: P=(1) Q=(1,2,5,6,7) -Ước lượng: Sau khi ước lượng và kiểm tra nhiều mô hình tôi thấy mô hình ARIMA( 0,1,7) là phù hợp nhất Bảng kết quả ước lượng: _Dự Báo : Tại ... forecast BÁO CÁO MÔN KINH TẾ LƯỢNG ỨNG DỤNG MÔ HÌNH ARIMA TRONG DỰ BÁO GIÁ DẦU THÔ THẾ GIỚI Sinh viên thực hiện: Đặng Thị Thu Hiền Mã sinh viên: CQ500927 Giới thiệu về mô hình Arima Trong ... chúng tôi đề xuất sử dụng mô hình ARIMA và phương pháp Box-jenkins để dự báo giá dầu thô thế giới trong ngắn hạn căn cứ vào chuỗi dữ liệu quá khứ. II. Xây dựng mô hình Arima cho giá dầu thô thế...

Ngày tải lên: 25/03/2014, 09:32

... dụng dự báo giá cá sông tại Tp. HCM 5 MÔ HÌNH ARIMA MÔ HÌNH ARIMA  Tính dừng (Stationary)  Tính mùa vụ (Seasonality)  Nguyên lý Box-Jenkin  Nhận dạng mô hình ARIMA  Xác định thông số mô hình ... src="data:image/png;base64,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 3 GIỚI THIỆU GIỚI THIỆU  Mô hình nhân quả  Mô hình chuỗi thời gian Hai loại mô hình dự báo chính: 2 NỘI DUNGNỘI DUNG  Giới thiệu xây dựng Mô Hình ARIMA (Auto-Regressive Integrated ... 10 MÔ HÌNH ARIMA MÔ HÌNH ARIMA Theo Box- Jenkin mọi quá trình ngẫu nhiên có tính dừng đều có thể biểu diễn bằng mô hình ARIMA 4  Đối với các chuỗi thời gian  ARIMA thường được sử dụng để dự...

Ngày tải lên: 02/04/2014, 21:59

... thể kết luận mô hình không có ảnh hưởng của ARCH 68 Hình 28:Kết quả ước lượng của mô hình AR(1) Hình 29: Các chỉ tiêu dự báo của mô hình AR(1) trains_times ... Dax (Đức),Train Times (Singapore) 40 1.5.3.Các chỉ tiêu dự báo và giá trị dự báo: Hình 11:Các chỉ tiêu dự báo của mô hình AR(2) stt Vnindex Vnindexfar2 143 315.724 354.1707502 ... W W − = − 2.2 .Mô hình ARIMA Theo Box-Jenkins một quá trình ngẫu nhiên có tính dừng để có thể biểu diễn bằng mô hình Tự hồi quy kết hợp Trung bình trượt ARIMA 2.2.1 .Mô hình tự hồi quy bậc...

Ngày tải lên: 12/04/2014, 23:28

... 04/2012 56 Hình 3.2: Biểu đồ tương quan chuỗi WQI 67 Hình 3.3: Ước lượng mô hình ARIMA (1, 0, 1) 68 Hình 3.4: Kiểm định mô hình ARIMA (1, 0, 1) 69 Hình 3.5: Kết quả dự báo của mô hình 70 ... được quan trắc, lập mô hình dự báo bằng mô hình ARIMA. Từ đó đưa ra những dự báo diễn biến chất lượng nước sông. Đánh giá khả năng áp dụng của mô hình ARIMA cho bài toán dự báo chất lượng nước. ... THUYẾT MÔ HÌNH ĐƯỢC SỬ DỤNG 2.1. Giới thiệu mô hình ARIMA Mô hình ARIMA (AutoRegressive Intergrate Moving Avarage) do Box – Jenkins đề nghị vào năm 1976, dựa trên mô hình tự hồi quy AR và mô hình...

Ngày tải lên: 26/04/2014, 12:35

... chọn là Mô hình 1. 2.4. Dự báo bằng mô hình ARIMA Mô hình ARIMA (0,1,1)(0,1,1) 12 tt uLLYLL )1)(1()1)(1( 12 11 12 Θ−−=−− θ (2.8) Tuy nhiên, để sử dụng một mô hình được nhận dạng cho dự báo, cần ... và dự báo lạm phát. Mục đích của bài viết này nhằm ứng dụng mô hình ARIMA với phương pháp Box-Jenkins để dự báo lạm phát ở Việt Nam. George Box và Gwilym Jenkins (1976) đã nghiên cứu mô hình ARIMA (Autoregressive ... hoá mô hình không. Bảng 2.2. Kết quả các thông số kiểm định Model Obs Chi-Square ll(model) df AIC BIC Mô hình 1 119 109.42 -85.63 4 169.26 180.38 Mô hình 2 119 45.94 -109.78 4 227.55 238.67 Mô hình...

Ngày tải lên: 14/07/2014, 00:20

Ngày tải lên: 18/07/2014, 17:26

Ngày tải lên: 24/07/2014, 01:20

Ngày tải lên: 08/08/2014, 04:21

... số, mô hình sẽ đưa ra dự báo cho ngày tiếp theo. 3.1.3. Thiết kế mô hình ARIMA cho dữ liệu Việc thiết kế thành công mô hình ARIMA phụ thuộc vào sự hiểu biết rõ ràng về vấn đề, về mô hình, ... 3.1.3.6 Dự báo ngắn hạn mô hình Dựa vào mô hình được chọn là tốt nhất, với dữ liệu quá khứ tới thời điểm t, ta sử dụng để dự báo cho thời điểm kế tiếp t+1. 3.2. Áp dụng Ứng dụng mô hình ARIMA ... trong quá trình xây dựng mô hình ARIMA. Giới thiệu sơ bộ về phần mềm ứng dụng Eviews 5.1 phục vụ cho bài toán dự báo bằng mô hình ARIMA. MỞ ĐẦU Bài toán dự báo tài chính ngày càng...

Ngày tải lên: 26/04/2013, 15:43