174 research outputs found
Analysis of Retailer Inventory and Financial Performance
No full textThis paper attempts to recreate the regression model originally presented in Kesavan and Mani (2013) to analyze the relationship between abnormal inventory growth (AIG) and one-year-ahead earnings per share (EPS) for U.S. public retailers. In addition, this paper aims to build upon Kesavan and Mani (2013)’s findings by applying the model to recent data in order to test whether results vary as a function of different macroeconomic conditions. Unlike Kesavan and Mani (2013), I do not find a statistically significant relationship between AIG and future EPS for the years 2004-2009. However, when applying the same model to data from 2013 to 2018, I find a significant, inverted-U relationship between the two variables. These findings suggest that abnormal inventory growth is impacted by macroeconomic factors that encourage retailers to accumulate excess inventory. Furthermore, I find that excess inventories have a larger negative impact on future earnings than insufficient inventories, implying that retailers should prioritize strategies that prevent bloated inventory levels above those that lead to decreased service level.Grossman School of Busines
Architecture validation of VFP control for the WiNC2R platform
No full textA Cognitive Radio processing requires intelligent transceiver which can be easily programmed and reconfigured dynamically to support multiple protocols. The Winlab Network Centric Cognitive Radio (WiNC2R) platform is based on the concept of Virtual Flow Pipelining Paradigm. WiNC2R can support per packet protocol adaption through the reconfiguration of function sequencing. Since WiNC2R platform can be programmed by adding additional functions in software, and flow sequencing reprogramming architecturally supported in hardware, it can easily support future protocols. The latest version of WiNC2R has advanced shared VFP control unit, cluster based SoC architecture with all the processing engines in an 802.11a like OFDM transmitter flow. It is very important to characterize the VFP overhead with the realistic protocol processing examples to understand the performance and cost penalties of added flexibility, and establish the base for the comparison with Software Defined Radio approach. The performance analysis of the VFP will give detailed insight about the various latencies involved in the VFP processing. VFP Architecture is validated to see that the current implementation does meet the requirements of the WiNC2R platform. This performance analysis will help in characterizing VFP overhead under varying throughput requirements. Architectural validation of VFP will characterize certain parameters of the system programming, like reschedule period, guard time, etc.M.S.Includes bibliographical referencesby Akshay Jo
Content-based image retrieval of digitized histopathology via boosted spectral embedding (BoSE)
No full textContent-based image retrieval (CBIR) systems allow for retrieval of images from a database that are similar in visual content to a query image. This is particularly useful in scenarios such as digital pathology, where text-based descriptors alone might be inadequate to accurately describe image content. By representing images via a set of quantitative image descriptors, the similarity between a query image with respect to archived, annotated images in a database can be computed and the most similar images retrieved. Recently, non-linear dimensionality reduction (NLDR) methods have become popular for embedding high dimensional data into a reduced dimensional space while preserving local object adjacencies, thereby allowing for object similarity to be determined more accurately in the reduced dimensional space. However, most dimensionality reduction (DR) methods implicitly assume, in computing the reduced dimensional representation, that all features are equally important. Erroneous or noisy features could potentially result in dissimilar images being mapped close to each other in the reduced embedding space. In this work we present Boosted Spectral Embedding (BoSE), a variant of the traditional Spectral Embedding (SE) NLDR method, which unlike SE utilizes a boosted distance metric (BDM) to selectively weight individual features to subsequently map the data into a reduced dimensional space. In this work BoSE is evaluated against SE (which employs equal feature weighting) in the context of CBIR of digitized prostate and breast cancer histopathology images. Across 154 hematoxylin and eosin (H&E) stained histopathology images corresponding to benign and malignant prostate cancer biopsy images, low and high grade ER+ breast cancer studies, and HER2+ breast cancer H&E images, BoSE outperformed SE both in terms of CBIR-based (area under the precision recall curve) and classifier-based (classification accuracy) performance measures. Consistent trends were observed when embedding the data into spaces with different dimensions. Our results suggest that BoSE could serve as an important tool for CBIR and classification of high dimensional biomedical data.M.S.Includes bibliographical referencesIncludes vitaby Akshay Sridha
Weighted K-nearest neighbor algorithm as an object localization technique using passive RFID tags
No full textTechnologies using identification by radio frequencies (RFID) are experiencing rapid development and healthcare is a major application area benefiting from it. Highly pervasive RFID enables remote identification, tracking and localization of the medical staff, patients, medications and equipment, thus increasing safety, optimizing in real-time management and providing support for new ambient-intelligent services. This thesis describes and evaluates an algorithm that enables object localization and tracking using passive RFID tags. This thesis also describes scenarios of how this technology can be used as a part of building a smart trauma resuscitation room by tracking the equipments. The main contribution of this thesis is the adaptation of the Weighted K-Nearest Neighbor Algorithm as a localization technique to track objects in a confined and crowded space by using passive RFID tags. The input parameter to the algorithm is the received signal strength indicator (RSSI), which gives a measure of back-scattered radio frequencies from passive tags. While using RFID technology special attention has to be given to the placement of antennas to get the optimum result. Therefore, we analyzed various antenna placement configurations with mean error and error consistency as the two performance parameters. The detection of multiple tags and human occlusion are two major concerns while tracking tags in a confined space with many team members collaborating on solving a problem. The RF signal can be interrupted by people walking around randomly and holding multiple (tagged) instruments at the same time. While the algorithm worked fine when tracking multiple tags, we had to modify the experimental set-up and attach an antenna onto the ceiling (which we call a vertical antenna), so that even if all the wall antennas are blocked we get at least one input parameter to base our localization decision on. We evaluated the algorithm for different combinations of configurations and number of neighbors, and achieved the following results. The best results were obtained for the 3 antennae (placed orthogonally) configuration considering the 4 nearest neighbors wherein a mean error rate of 15% of the maximum possible error was achieved under ideal conditions. We tested the algorithm for different human occlusion scenarios i.e. blocking 1 or 2 wall antennas, standing in random positions and then roaming in the field area randomly. The mean error rate for the standing scenario was measured as 20% of the maximum possible error and 18% in the case of roaming configuration. The error was found to be consistently within our defined maximum error for 100% of the recorded readings. The results obtained were found to be satisfactory for our application where, more than the exact location of the object, knowing whether the object is within a particular region is good enough for the users to know what task is being carried out in the trauma bay. Also the algorithm holds good in an indoor environment having a lot of factors and materials which affect the RF signal disrupting accurate calculation of the location co-ordinates. The algorithm does not require extensive data collection prior to implementation which makes it easily deployable in any environment. Apart from the problems mentioned there are some other factors like materials on which the tags are attached and orientation of tags which were found to be potential hindrances for accurate localization. Acceptable solutions to these problems form a part of our future work.M.S.Includes bibliographical referencesby Akshay Shett
Strategic network design for reverse logistics
No full textReverse logistics (RL) is gaining importance both in terms of research and industrial applications due to the thrust by the government, environmentalists and consumers to conserve resources. The issue in RL is to take back the used products, so that the product or its parts are appropriately disposed, recycled, reused, or remanufactured. Most of the researchers do not recognize the complexity in handling returns of modular products. Also, most of the models proposed in the literature are not suitable for multi-product configurations. Thus, a deterministic optimization model is proposed to design a network for recovery of products. The model considers supply of return products, demand for the remanufactured products, supply of reusable modules and materials in the secondary market and is suitable for multi-product configurations. The network also considers suppliers of new modules to be used during the final assembly of the product. However, in practice, the return of used products and the demand for remanufactured products is uncertain. Hence, the base model is modified to incorporate stochastic supply and demand scenarios. The proposed networks are analyzed by using simulated data to show the applicability of the model in real life situations. Scenario analyses are performed on the model to depict real life situations. The proposed network can assist in strategic decision making in RL networks.MASTER OF ENGINEERING (MAE
A Global Analysis on Microgrids through the PESTEL Framework
No full textMicrogrids enable distribution of electricity with higher shares of variable renewables, higher power quality, greater reliability and higher efficiency. There are a large number of factors in addition to the technology, which affect their shift towards market competitiveness and widespread adoption. The PESTEL framework, covering Political, Economic, Social, Technical, Environmental and Legislative factors, is used to identify and describe the drivers and barriers for microgrid development at the global level. The framework enables a broader approach to describe potential for microgrid applications. The results aim to provide engineers, project developers and microgrid specialists with an overview of the prospects for microgrid deployment.Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Energy Technolog
How does trade impact agricultural productivity?
No full textThe student, Akshay Pandit, submitted this Thesis for approval on 2020-07-22 at 15:29.This Thesis was approved for publication on 2020-07-23 at 10:50.DSpace SAF Submission Ingestion Package generated from Vireo submission #15729 on 2020-10-02 at 15:34:07Made available in DSpace on 2020-10-07T22:44:48Z (GMT). No. of bitstreams: 2
PANDIT-THESIS-2020.pdf: 10275210 bytes, checksum: bdf6f32a4714aaadf246aa27560ec60f (MD5)
LICENSE.txt: 4210 bytes, checksum: ad7b57595833966ecb91704e689e58e5 (MD5)
Previous issue date: 2020-07-23"Agricultural production has faced increased demands over the last half century from an expanding economy and population. We live in a globalized world, in which agriculture is deeply intertwined in international markets and trade. In this paper, we address the overarching research question: ""What is the impact of trade on agricultural productivity?''. To this end, we present a comprehensive statistical and econometric analysis on the relationship between international trade and agricultural production. We use national-scale data on crop yield, area harvested, production, and trade for the last half century (1961-2016) from the Food and Agricultural Organization of the United Nations. We introduce novel weighting and decomposition analyses to explore the relationship between trade and crop productivity. To determine the causal impact of trade on agriculture we implement instrumental variable (IV) econometric methods. We find that trade has led to an increase in global agricultural productivity over time (e.g. through increased productivity, the intensive margin). Global productivity gains have accrued primarily through the participation of more countries in global trade (e.g. expanding the area of contribution, the extensive margin). Additionally, we find that trade has enabled global crop consumption to increase. These findings indicate that trade openness leads to greater productivity in agriculture in general. This work highlights that trade can help to achieve productivity gains in agriculture and potentially help the world to address remaining yield gaps."Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2022-08-01The student, Akshay Pandit, accepted the attached license on 2020-07-22 at 15:28.Embargo set by: Seth Robbins for item 116267
Lift date: 2022-10-07T22:44:53Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemAuthor requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Onl
Behavioral data collection and simulation
No full textThesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2017.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.Cataloged from student-submitted PDF version of thesis.Includes bibliographical references (page 45).On-demand ridesharing services, such as Uber and Lyft, and autonomous vehicles are significantly changing the landscape of transportation and mobility. In light of these disruptions, we aim to determine consumer preferences with regards to transportation and use this data to simulate and analyze the urban effects of smart mobility solutions. We collect behavioral data using Future Mobility Sensing (FMS), a smartphone and prompted-recall-based integrated activity-travel survey, and create simulations using the data with SimMobility, a simulation platform that integrates various mobility-sensitive behavioral models with state-of-the-art scalable simulators to predict the impact of mobility demands on transportation networks, intelligent transportation services, and vehicular emissions. Enhancing these projects with on-demand preferences, individual patterns, and incentives as inputs, we aim to simulate and analyze a wide range of viable smart mobility solutions.by Akshay Padmanabha.M. Eng
A framework to search for machine learning pipelines
No full textThesis: M. Eng. in Computer Science, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2018.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.Cataloged from student-submitted PDF version of thesis.Includes bibliographical references (page 81).In this thesis, we present DeepMining, a framework to search for machine learning pipelines. The high-level goal of DeepMining is to solve the pipeline search problem: given a problem and a dataset, find the pipeline best-suited to solve that problem. The DeepMining platform serves as a testbed for developers to experiment with different methods of computing and evaluating machine learning pipelines. Specifically, developers have autonomy over how to evaluate different configurations in parallel, score a pipeline given a dataset and hyperparameter configuration, and efficiently search over the pipeline space. DeepMining was designed with modularity and extensibility in mind: developers can easily implement new search algorithms, scoring functions, and computation frameworks. At the same time, users can switch between these modules with minimal effort.by Akshay Ravikumar.M. Eng. in Computer Scienc
Logistic Network Analysis for Stochastic Supply and Demand of Used Products
No full textReverse logistics (RL) is concerned with reverse flow of the products from consumer to the original equipment manufacturers. The development of the network for collection, storage, inspection, remanufacturing and redistribution are important RL activities that need to be carefully planned.RL is characterized by uncertainty in terms of the receipt of the used products from the consumer. A good returned product can be reused or remanufactured with lower inputs. However, a defective product can either be disposed or recycled. On the other hand, it is also difficult to predict as to what the demand for the remanufactured or reusable products would be. Thus RL has uncertainty as to receipt (of used products) and their processing. Therefore, while designing a network, companies have to consider not only the allocation of capacities for warehouse, transportation but also the stochastic nature of supply and demand. Remanufacturing not only requires the used parts but it also needs to be supplemented by new modules.In this work, we propose a model that considers stochastic supply of return products and stochastic demand for the remanufactured products. The network also considers the sale of reusable modules and materials in the secondary market and the requirements for new modules for remanufacturing in case of higher demand for the remanufactured products. The attempt is to establish a network incorporating the above randomness for strategic decision involving huge capital investments. The network is illustrated by a numerical study
- …
