The use of technology and information as multimedia learning has a very strategic function. Media is seen as one of the factors that can improve the learning process because the media has strategic roles and functions that can directly or indirectly affect the motivation, interests, and attention of students. The function of multimedia in learning is to overcome the limitations of students' experiences, be able to transcend classroom boundaries, enable direct interaction between students and their environment, produce uniform observations, instill the basic concepts that are true, concrete, and realistic. Multimedia can arouse new desires and interests and arouse students' motivation to learn and provide an integral or comprehensive experience from the concrete to the abstract. The advantages of using multimedia in learning include being able to improve students' ability to understand an abstract concept more easily, in addition to the use of computer media in the form of multimedia can give a positive impression to the teacher because it can help teachers explain the contents of the lesson to students, streamline time, and increase students' motivation in learning.

No abstract

Digital image watermarking is frequently used for many purposes, such as image authentication, fingerprinting, copyright protection, and tamper proofing. Imperceptibility and robustness are the watermark requirements of good watermarks. In this paper, we propose the Fast Walsh Hadamard transform (FWHT) combined with the Discrete Cosine Transform (DCT) as a new image watermarking scheme. The FWHT reorders the high-to-Iow sequence components contained in the signal. This scheme produces high perceptual transparency of the embedded watermark.Experimental results show that the proposed scheme has good visual perception and is robust against attacks.

The stock price prediction or forecasting is one of the important information in the stock market. Investors need stock price information before buying shares, while shareholders need stock price information for selling shares. There are many stock price prediction techniques have been proposed in the time series analysis. One of the simplest and powerful techniques is singular spectrum analysis, which works on the time series decomposition that is constructed using sub-series or window of the initial time series. However, choosing the exact window length is not easy because it depends on the time series characteristics. In this paper, the Hadamard transform of time series is proposed as an alternative technique to choose the window length in time series embedding. Technically, the length of the window for time series embedding is determined directly based on the size of the Hadamard spectrum. The experimental results show that the proposed method not only facilitates the determination of window length in time series embedding but can also improve the performance of the standard singular spectrum analysis method. The error rates of the proposed and the baseline methods (the standard singular spectrum analysis and the standard singular spectrum analysis with minimum description length) are 0.0088, 0.0194, and 0.1441, respectively.