Browsing by Author "Ahmed, Saif"

Now showing 1 - 7 of 7

A voice signal based gender prediction model using random forest classifier
(BRAC University, 2018-04) Ahmed, Saif; Hossain, Sajjad; Chowdhury, Gazala; Mehnaz, Maliha; Uddin, Jia
In the proposed model, Classification and Regression Tree (CART) was used as a classifier to classify gender using four different algorithms which were tested with changing dataset frames, layer sizes and samples to get best options for our model. We had to tune our dataset with Principal Component Analyzer(PCA) which improved the accuracy rate a bit and also worked along with the algorithms. The intelligible idea of voiceprints and human-computer interaction gave us the motivation to predict gender by using different proposed classifiers that we are using in our model .Besides the overall efficiency and outcome of human-computer interaction gave us the inspiration to select this model for our thesis paper. In this existing system there are quite a lot of problem that arose while dealing with our proposed model those are over fitting of the dataset, having different layer sizes, number of decision tree and most importantly solving the hidden layer sizes. We did successfully solved most of the problems by running five different algorithms on our model which are Decision Tree Classifier, Logistic Regression, Support Vector Machine (SVM) , Multi-Layer Perceptron Classifier (MLP) and Random Forest (RF) Classifier. To use the total dataset on this algorithm we used 75% training and 25% testing of the total dataset. Due to different layers we had different accuracy result for each of the algorithms. The worst accuracy result was given by Multi-Layer Perceptron (MLP) which was 75% in two implementations and the best accuracy result was given by Random Forest Classifier which was 97.34 % from our proposed model.
Analyzing intermediaries that determines purchase behavior of iPad users for digital products within iPadOS applications: consumer study for Lettre.app
(BRAC University, 2024-02) Ahmed, Saif; Tabassum, Takmilla
This report is a representation of my successful completion regarding a three-months internship program under CheeThee LLC. I served as a product marketer along with a few designs related responsibilities for product marketing sector. Lettre.app is an iPad exclusive product from a North American tech company named CheeThee LLC. The objective of this report is to recognize the evaluative process of the iPad users while attempting to purchase digital products or participating to in-app purchases within apps. The primary concern is to provide insights regarding the prices of digital offerings by Lettre.app and offer a comprehensive understanding towards iPad users. In order to accomplish the aim and effectively conduct the research project, I have been able to utilize a proper combination of primary and secondary sources. About primary sources, all the information and data were collected through the method of convenience sampling. In my study encompassing 45 participants, the predominant age group of iPad owners to be within the 21-25 age range, with a majority being students. A notable insight is majority of the users owns the complementary product named Apple Pencil that enhances the overall user behaviour toward iPad experience. Almost all users conduct in-app purchases, with their decisions influenced by moderately priced digital products or in-app services/products that cater to specific features. AppStore and other platforms alike contain user reviews that associates a vital role to guide users in terms of enhancing the decision-making process for paid applications. The analysis indicates about an enhanced post-purchase experience and additional value-added benefits which the are core motivators for in-app purchases and paid applications. According to the findings from this study, Lettre.app is suggested develop a targeted marketing campaigns and separate digital collectible items in order to attract students. In addition, incorporating event-based souvenirs and contests can strengthen user experiences. Implementing affordable stamp options as an effective penetrative pricing strategy might be attractive to the majority of users interested in making moderate investments in digital products.
Band structure study of graphene and transition metal dichalcogenides heterostructures
(BRAC University, 2017) Ahmed, Saif; Ahmed, Nafisa Nowshin; Nur, Juhayer An; Rahman, Md. Mosaddequr
According to Moore’s Law, it is predicted that the number of transistors –the building blocks of any modern electronic device– on integrated circuits per square inch will double every year. However studies show that the limit would be reached within the next three years as the size of Silicon based transistors, already reduced to a few tens of nanometers, can no longer be further reduced. This is not the only challenge for the semiconductor industry. Besides the transistor size, engineers are also not being able to decrease the minimum voltage required to turn on a device, making it tough for us to build even low power consuming devices. This has compelled us to focus on different alternatives. Graphene has turned out to be the wonder material in the scientific community recently, and its applications are innumerable. It is only an atomically thick 2 dimensional material with very high electron mobility. However, the only reason that acts as the barrier to fully utilize the great potential of this material is its missing band gap. This is when the role of transition metal dichalcogenides (TMDCs) comes into play, another atomically thick 2 dimensional set of materials, which, unlike Graphene, has direct band gap but low mobility i.e. characteristics that can compensate for those missing in Graphene. Scientists have recently combined Graphene with TMDCs to extract the best possible characteristics from both the materials in quest for a substitution for traditional Silicon based transistors in the industry and the results are promising; however, the low current on/off ratio of such devices is still a challenge that needs to be overcome. In our thesis we have simulated band structures of Graphene and different TMDCs like MoS2, MoT e2 etc. Additionally we have simulated band structures of Graphene interfaced with various TMDCs and also Graphene sandwiched between various TMDCs. Through literature review, we have also found out the spin orbit coupling – a phenomenon that allows us to control electricity through orientation of electrons’ polarization rather than the flow of charge – of Graphene and various TMDCs. We have then graphically analyzed our data and finally proposed that by combining the electronic band gap and the spin orbit coupling of a material simultaneously in the future, we may be able to create devices which will be much smaller in size and run with much lower power than current transistors and simultaneously have a high current on/off ratio.
Design of an Optimum Massive MIMO FSO System and Analysis of its Performance in Different Weather Conditions
(Research and Development Wing, MIST, 2022-12) Ahmed, Saif; Ahnaf Hasan, Syed Md.; Islam, A. K. M. Nazrul
This aim of this work is to propose a massive MIMO FSO system with optimum design parameters and analyze its performance in various turbulent weather conditions by simulating the model in Optisystem . In recent years, the need for high speed data transfer for real time communication has become a necessity. Implementation of massive MIMO system has already gained remarkable achievement using the conventional RF transmission system. Although it is very much efficient in its application, it has some problems, such as licensing restrictions, bandwidth depletion, fragmentation. In this respect, Free space optics (FSO) is an efficient method for high speed data transfer as it does not have above limitations. Different researches conducted using MIMO FSO models have shown promising results. The use of massive MIMO system (more than array of antennas) will enhance data rate further. In this research work, an optimum massive MIMO FSO system has been designed with array of antennas. The performance of this system is analyzed in different weather conditions and compared with that of system. The analysis of performance is done by power penalty and receiver sensitivity plots. A significant transmission range and an optimum performance is obtained by using this system at a data rate of Gbps. However, data transmission rate can be much more enhanced in this system at the cost of reduced performance and a shorter transmission range.
Detecting man in the middle attack implementing timebased verification method
(BRAC University, 2022-05) Topu, Syed Naziur Rahman; Ahmed, Saif; Shakil, Arif; Hossain, Muhammad Iqbal
Internet is one of the most important parts of our daily life. Network security is necessary to secure our internet network. For security purposes, we are developing different protocols, methods, and techniques. Still, those are not enough to secure the internet. Attackers are finding new loopholes to create different kinds of attacks. Man in the Middle attack is one of them. Attackers sit between a connected line and breach their privacy and steal their information. Many existing solutions have practical limitations, deployment costs as well as adversaries. To overcome this problem, we are suggesting a new program to detect the Man in the Middle attack at the beginning of the connection establishment and also alert the user about the attack with the attacker’s information
Different Touch - focusing on campaign
(BRAC University, 11/5/2017) Ahmed, Saif; Haque, Shamim Ehsanul
The internship report focuses on the experience I have gathered being an intern in the Marketing department of Different Touch which is the working for many renowned national and multinational brands. I have mostly looked into the marketing campaign that Different Touch worked on. Different Touch has always provided quality service since its inception in 1997. The company’s operations include a huge number of services which they deliver using different medium. I have shared a detailed insight about my work experience at Different Touch along with a detailed job description where I had to work with local government, conducting research through visits and periodic surveys, brainstorming, planning and implementing ideas on how to execute Different Touch's services in the best possible manner. I have written about the project I worked on such as “Street Drama”, “Pohela Boishakh Road Show”, “Documentary on Prime Minister Sheikh Hasina ’s Vow- No Children to be left on the Streets.” for Aparajeyo Bangladesh, an NGO working for safety of street children. I have worked on planning and positioning the campaign so as to make the message more effective and also to retain these campaigns in the consumers' mindsets. I have also mentioned about how the message and positioning of the message was done for these projects in an attempt to make the brands more retain these brands in the consumers' mindsets.
Power Consumption Analysis of UAVs with Varying Payloads for Next Generation Wireless Networks
(Research and Development Wing, MIST, 2025-12-30) Ahmed, Saif; Islam Emon, Nimul
Unmanned Aerial Vehicles (UAVs) are emerging as an important component for enhancing the coverage and capacity of next generation wireless networks. However, the use of UAVs is challenging because of crucial power management. The power management becomes more sophisticated for rotary-wing UAVs, the most commonly used ones, which require more energy as payload and flight actions increase. The current study investigates the energy consumption of UAVs with varied payloads during hovering and moving operations. The effects of payloads such as cameras, sensors, GPS modules, communication systems, and flight controllers on UAV power consumption are investigated in this study. It is very much necessary to comprehend these energy dynamics for optimal UAV design and maximizing operational effectiveness. Key findings show how payload configuration and flying conditions affect energy requirements. Two regression-based empirical mathematical models for hovering and moving conditions representing the relationship between the power consumption of UAV and its weight with payloads are proposed. This rigorous analysis of power consumption of UAV provides valuable insights into optimizing UAV design and operation in next-generation wireless networks. The findings of the work, including the proposed mathematical models, are expected to help researchers and engineers develop more energy- efficient UAVs, resulting in extended flight periods and improved operational reliability.