scholarly journals Optimizing the Frequency Capping: A Robust and Reliable Methodology to Define the Number of Ads to Maximize ROAS

2021 ◽  
Vol 11 (15) ◽  
pp. 6688
Author(s):  
Jesús Romero Leguina ◽  
Ángel Cuevas Rumin ◽  
Rubén Cuevas Rumin
Keyword(s):  
High Frequency ◽  
Large Scale ◽  
Low Frequency ◽  
Digital Marketing ◽  
Optimal Frequency ◽  
Novel Technique ◽  
Maximum Threshold

The goal of digital marketing is to connect advertisers with users that are interested in their products. This means serving ads to users, and it could lead to a user receiving hundreds of impressions of the same ad. Consequently, advertisers can define a maximum threshold to the number of impressions a user can receive, referred to as Frequency Cap. However, low frequency caps mean many users are not engaging with the advertiser. By contrast, with high frequency caps, users may receive many ads leading to annoyance and wasting budget. We build a robust and reliable methodology to define the number of ads that should be delivered to different users to maximize the ROAS and reduce the possibility that users get annoyed with the ads’ brand. The methodology uses a novel technique to find the optimal frequency capping based on the number of non-clicked impressions rather than the traditional number of received impressions. This methodology is validated using simulations and large-scale datasets obtained from real ad campaigns data. To sum up, our work proves that it is feasible to address the frequency capping optimization as a business problem, and we provide a framework that can be used to configure efficient frequency capping values.

Investigating Thunderstorm HF/VHF Radio Bursts with Weak Lower Frequency Radiation

2020 ◽  
Author(s):  
Ningyu Liu ◽  
Joseph Dwyer
Keyword(s):  
High Frequency ◽  
Large Scale ◽  
Strong Field ◽  
Large Fraction ◽  
Low Frequency ◽  
Very High Frequency ◽  
Lightning Initiation ◽  
Study Support ◽  
Volcanic Lightning ◽  
Frequency Radiation

<p>While the spectrum of lightning electromagnetic radiation is known to peak around 5-10 kHz in the very low frequency (VLF) range, intense high frequency/very high frequency (HF/VHF) radiation can be produced by various lightning related processes. In fact, thunderstorm narrow bipolar events (NBEs), which are capable of initiating lightning, are the most powerful HF/VHF sources in nature on Earth. But even for NBEs, the spectral intensity in HF/VHF is still many orders of magnitude weaker than that of lower frequencies (Liu et al., JGR, 124, https://doi.org/10.1029/2019JD030439, 2019). HF/VHF bursts with weak VLF signals, however, can also be produced by thunderstorms. These bursts may be related to the thunderstorm precursor events noted by Rison et al. (Nat. Commun., 7, 10721, 2016) and are also found to precede a large fraction of lightning initiation (Lyu et al., JGR, 124, 2994, 2019). They are also known as continual radio frequency (CRF) radiation associated with volcanic lightning (Behnke et. al., JGR, 123, 4157, 2018).</p><p> </p><p>In this talk, we report a theoretical and modeling study to investigate a physical mechanism for production of those HF/VHF bursts. The study is built on the theory developed recently concerning the radio emissions from an ensemble of streamers (Liu et al., 2019). We find an ensemble of streamer discharges that develop in random directions can produce HF/VHF radiation with intensity comparable to those all developing in a single direction, but the VLF intensity is many orders of magnitude weaker. The results of our study support the conclusions of Behnke et. al (2018) that CRF is produced in the absence of large-scale electric field, it results in insignificant charge transfer, and it is caused by streamers. In the context of the HF/VHF bursts preceding lightning initiation (Lyu et. al, 2019), our results imply that highly localized strong field regions exist in thunderstorms and streamers take place in those regions, which somehow precondition the medium for lightning initiation.</p>

scholarly journals Drag forces on a bed particle in open-channel flow: effects of pressure spatial fluctuations and very-large-scale motions

2019 ◽  
Vol 863 ◽  
pp. 494-512 ◽  
Author(s):  
S. M. Cameron ◽  
V. I. Nikora ◽  
I. Marusic
Keyword(s):  
Channel Flow ◽  
Drag Force ◽  
High Frequency ◽  
Large Scale ◽  
Open Channel ◽  
Low Frequency ◽  
Open Channel Flow ◽  
Stereoscopic Particle Image Velocimetry ◽  
Force Fluctuations ◽  
Drag Forces

The fluctuating drag forces acting on spherical roughness elements comprising the bed of an open-channel flow have been recorded along with synchronous measurements of the surrounding velocity field using stereoscopic particle image velocimetry. The protrusion of the target particle, equipped with a force sensor, was systematically varied between zero and one-half diameter relative to the hexagonally packed adjacent spheres. Premultiplied spectra of drag force fluctuations were found to have bimodal shapes with a low-frequency (${\approx}0.5~\text{Hz}$) peak corresponding to the presence of very-large-scale motions (VLSMs) in the turbulent flow. The high-frequency ($\gtrapprox 4~\text{Hz}$) region of the drag force spectra cannot be explained by velocity time series extracted from points around the particle, but instead appears to be dominated by the action of pressure gradients in the overlying flow field. For small particle protrusions, this high-frequency region contributes a majority of the drag force variance, while the relative importance of the low-frequency drag force fluctuations increases with increasing protrusion. The amplitude of high-frequency drag force fluctuations is modulated by the VLSMs irrespective of particle protrusion. These results provide some insight into the mechanics of bed particle stability and indicate that the optimum conditions for particle entrainment may occur when a low-pressure region embedded in the high-velocity portion of a VLSM overlays a particle.

scholarly journals What Caused The Increase of Tropical Cyclones In The Western North Pacific During The Period of 2011-2020?

2021 ◽  
Author(s):  
Haili Wang ◽  
Chunzai Wang
Keyword(s):  
Tropical Cyclones ◽  
North Pacific ◽  
High Frequency ◽  
Western North Pacific ◽  
Large Scale ◽  
Low Frequency ◽  
Vertical Wind Shear ◽  
Significant Negative Correlation ◽  
Relative Vorticity ◽  
The Pacific

Abstract Based on satellite era data after 1979, we find that the tropical cyclone (TC) variations in the Western North Pacific (WNP) can be divided into three-periods: a high-frequency period from 1979-1997 (P1), a low-frequency period from 1998-2010 (P2), and a high-frequency period from 2011-2020 (P3). Previous studies have focused on WNP TC activity during P1 and P2. Here we use observational data to study the WNP TC variation and its possible mechanisms during P3. Compared with P2, more TCs during P3 are due to the large-scale atmospheric environmental conditions of positive relative vorticity, negative vertical velocity and weak vertical wind shear. Warmer SST is found during P3, which is favorable for TC genesis. The correlation between the WNP TC frequency and SST shows a significant positive correlation around the equator and a significant negative correlation around 36°N, which is similar to the warm phase of the Pacific Decadal Oscillation (PDO). The correlation coefficient between the PDO and TC frequency is 0.71, significant at 99% confidence level. The results indicate that the increase of the WNP TC frequency during 2011-2020 is associated with the phase transition of the PDO and warmer SST. Therefore, more attention should be given to the warmer SST and PDO phase when predicting WNP TC activity.

scholarly journals The budget of local available potential energy of low-frequency eddies in Northern Hemispheric winter

2020 ◽  
pp. 1-46
Author(s):  
Xianglin Dai ◽  
Yang Zhang ◽  
Xiu-Qun Yang
Keyword(s):  
Potential Energy ◽  
Energy Budget ◽  
High Frequency ◽  
Large Scale ◽  
Vertical Structure ◽  
Low Frequency ◽  
Available Potential Energy ◽  
Background Temperature ◽  
Northern Hemispheric ◽  
Climate Events

AbstractLow-frequency (LF) transient eddies (intra-seasonal eddies with time scales longer than 10 days) is increasingly found important in large-scale atmospheric circulations, high-impact climate events and subseasonal-to-seasonal forecast. In this study, features and the maintenance of available potential energy of LF eddies (LF EAPE), which denotes LF temperature fluctuations, have been investigated. Our study shows that wintertime LF EAPE, with greater amplitude than that of the extensively studied high-frequency (HF) eddies, exhibits distinct horizontal and vertical structures. Different from HF eddies, whose action centers are over midlatitude oceans, the LF EAPE is most active in the continents in midlatitude, as well as the subpolar region with shallower vertical structure. By diagnosing the derived energy budget of LF EAPE, we find that, with the strong background temperature gradient in mid- and high-latitude continents (e.g. coast regions along Greenland-Barents-Kara sea), baroclinic generation is the major source of LF EAPE. The generated LF EAPE in the subpolar region is transported downstream and southward to midlatitude continents via background flow. The generated LF EAPE is also dissipated by HF eddies, damped by diabatic effect and converted to LF EKE via vertical motions. The above energy budget, together with the barotropic dynamics revealed by previousworks, suggests multiple energy sources thus complicated dynamics of LF variabilities.

On BFSA Collision Resolution in LF, HF, and UHF RFID Networks

2015 ◽  
Vol 4 (2) ◽  
pp. 44-55
Author(s):  
Varun Bhogal ◽  
Zornitza Genova Prodanoff ◽  
Sanjay P. Ahuja ◽  
Kenneth Martin
Keyword(s):  
Radio Frequency Identification ◽  
High Frequency ◽  
Large Scale ◽  
Low Frequency ◽  
Rfid Tags ◽  
Uhf Rfid ◽  
Ceteris Paribus ◽  
Frequency Identification ◽  
Major Factors ◽  
The Cost

RFID (radio frequency identification) technology has gained popularity in a number of applications. Decreased cost of hardware components along with wide adoption of international RFID standards have led to the rise of this technology. One of the major factors associated with the implementation of RFID infrastructure is the cost of tags. RFID tags operating in the low frequency spectrum are widely used because they are the least expensive, but have a small implementation range. This paper presents an analysis of RFID performance across low frequency (LF), high frequency (HF), and ultra-high frequency (UHF) environments. The authors' evaluation is theoretical, using a passive-tag BFSA based simulation model that assumes 10 to 1,500 tags per reader and is created with OPNET Modeler 17. Ceteris paribus, the authors' results indicate that total census delay is lowest for UHF tags, while network throughput performance of LF tags is highest for large scale implementations of hundreds of tags in reader's range. A statistical analysis has been conducted on the findings for the three different sets.

Wave Propagation in Continuous Sea Ice: An Experimental Perspective

2020 ◽  
Author(s):  
Giulio Passerotti ◽  
Alberto Alberello ◽  
Azam Dolatshah ◽  
Luke Bennetts ◽  
Otto Puolakka ◽  
...  
Keyword(s):  
Sea Ice ◽  
Wave Energy ◽  
High Frequency ◽  
Large Scale ◽  
Low Frequency ◽  
Ocean Waves ◽  
Continuous Model ◽  
Ice Cover ◽  
Irregular Waves ◽  
Ice Melt

Abstract Ocean waves penetrate hundreds of kilometres into the ice-covered ocean. Waves fracture the level ice into small floes, herd floes, introduce warm water and overwash the floes, accelerating ice melt and causing collisions, which concurrently erodes the floes and influences the large-scale deformation. Concomitantly, interactions between waves and the sea ice cause wave energy to reduce with distance travelled into the ice cover, attenuating wave driven effects. Here a pilot experiment in the ice tank at Aalto University (Finland) is presented to discuss how the properties of irregular small amplitude (linear) waves change as they propagate through continuous model sea ice. Irregular waves with a JONSWAP spectral shape were mechanically generated with a very low initial wave steepness to avoid ice break up and maintain a consistent continuous ice cover throughout the experiments. Observations show an exponential attenuation of wave energy with distance. High frequency components attenuated more rapidly than the low frequency counterparts, in agreement with a frequency-cubed power-law. The more effective attenuation in the high frequency range induced a substantial downshift of the spectral peak, stretching the dominant wave component as it propagates in ice.

Ritual Frequency, Emotionality, and Modes of Religiosity

2021 ◽  
pp. 53-81
Author(s):  
Harvey Whitehouse
Keyword(s):  
High Frequency ◽  
Large Scale ◽  
Low Frequency ◽  
New Guinea ◽  
Small Scale ◽  
Complex Societies ◽  
High Arousal ◽  
The World ◽  
Hierarchical Groups ◽  
To Come

Collective rituals tend to come in two kinds: frequently performed but relatively lowkey; rarely enacted but emotionally intense. According to the theory of modes of religiosity, high-frequency but low-arousal rituals produce large-scale hierarchical groups (the doctrinal mode), while low-frequency but high-arousal rituals produce small-scale highly cohesive groups (the imagistic mode). This chapter describes how that theory was first developed while carrying out fieldwork in the New Guinea rainforest. But then the author realized it could help to explain how groups throughout the world take shape and spread, and it could also help to explain how complex societies evolved.

scholarly journals Investigation of High-Frequency Internal Wave Interactions with an Enveloped Inertia Wave

2012 ◽  
Vol 2012 ◽  
pp. 1-12
Author(s):  
B. Casaday ◽  
J. Crockett
Keyword(s):  
Internal Wave ◽  
High Frequency ◽  
Wave Breaking ◽  
Large Scale ◽  
Critical Level ◽  
Low Frequency ◽  
Small Scale ◽  
Wave Interactions ◽  
Frequency Wave ◽  
Action Density

Using ray theory, we explore the effect an envelope function has on high-frequency, small-scale internal wave propagation through a low-frequency, large-scale inertia wave. Two principal interactions, internal waves propagating through an infinite inertia wavetrain and through an enveloped inertia wave, are investigated. For the first interaction, the total frequency of the high-frequency wave is conserved but is not for the latter. This deviance is measured and results of waves propagating in the same direction show the interaction with an inertia wave envelope results in a higher probability of reaching that Jones' critical level and a reduced probability of turning points, which is a better approximation of outcomes experienced by expected real atmospheric interactions. In addition, an increase in wave action density and wave steepness is observed, relative to an interaction with an infinite wavetrain, possibly leading to enhanced wave breaking.

scholarly journals Complexity Changes in the US and China’s Stock Markets: Differences, Causes, and Wider Social Implications

Entropy ◽  
2020 ◽  
Vol 22 (1) ◽  
pp. 75
Author(s):  
Jianbo Gao ◽  
Yunfei Hou ◽  
Fangli Fan ◽  
Feiyan Liu
Keyword(s):  
Stock Market ◽  
Time Scales ◽  
Stock Markets ◽  
High Frequency ◽  
Large Scale ◽  
Low Frequency ◽  
Stock Index ◽  
Chinese Market ◽  
Index Data ◽  
The Us

How different are the emerging and the well-developed stock markets in terms of efficiency? To gain insights into this question, we compared an important emerging market, the Chinese stock market, and the largest and the most developed market, the US stock market. Specifically, we computed the Lempel–Ziv complexity (LZ) and the permutation entropy (PE) from two composite stock indices, the Shanghai stock exchange composite index (SSE) and the Dow Jones industrial average (DJIA), for both low-frequency (daily) and high-frequency (minute-to-minute)stock index data. We found that the US market is basically fully random and consistent with efficient market hypothesis (EMH), irrespective of whether low- or high-frequency stock index data are used. The Chinese market is also largely consistent with the EMH when low-frequency data are used. However, a completely different picture emerges when the high-frequency stock index data are used, irrespective of whether the LZ or PE is computed. In particular, the PE decreases substantially in two significant time windows, each encompassing a rapid market rise and then a few gigantic stock crashes. To gain further insights into the causes of the difference in the complexity changes in the two markets, we computed the Hurst parameter H from the high-frequency stock index data of the two markets and examined their temporal variations. We found that in stark contrast with the US market, whose H is always close to 1/2, which indicates fully random behavior, for the Chinese market, H deviates from 1/2 significantly for time scales up to about 10 min within a day, and varies systemically similar to the PE for time scales from about 10 min to a day. This opens the door for large-scale collective behavior to occur in the Chinese market, including herding behavior and large-scale manipulation as a result of inside information.

scholarly journals Context-aware Retrieval-based Deep Commit Message Generation

2021 ◽  
Vol 30 (4) ◽  
pp. 1-30
Author(s):  
Haoye Wang ◽  
Xin Xia ◽  
David Lo ◽  
Qiang He ◽  
Xinyu Wang ◽  
...  
Keyword(s):  
Machine Translation ◽  
High Frequency ◽  
Large Scale ◽  
Low Frequency ◽  
Ground Truth ◽  
Poor Quality ◽  
Context Aware ◽  
Training Set ◽  
Neural Machine Translation ◽  
Large Scale Dataset

Commit messages recorded in version control systems contain valuable information for software development, maintenance, and comprehension. Unfortunately, developers often commit code with empty or poor quality commit messages. To address this issue, several studies have proposed approaches to generate commit messages from commit diffs . Recent studies make use of neural machine translation algorithms to try and translate git diffs into commit messages and have achieved some promising results. However, these learning-based methods tend to generate high-frequency words but ignore low-frequency ones. In addition, they suffer from exposure bias issues, which leads to a gap between training phase and testing phase. In this article, we propose CoRec to address the above two limitations. Specifically, we first train a context-aware encoder-decoder model that randomly selects the previous output of the decoder or the embedding vector of a ground truth word as context to make the model gradually aware of previous alignment choices. Given a diff for testing, the trained model is reused to retrieve the most similar diff from the training set. Finally, we use the retrieval diff to guide the probability distribution for the final generated vocabulary. Our method combines the advantages of both information retrieval and neural machine translation. We evaluate CoRec on a dataset from Liu et al. and a large-scale dataset crawled from 10K popular Java repositories in Github. Our experimental results show that CoRec significantly outperforms the state-of-the-art method NNGen by 19% on average in terms of BLEU.

Sign in / Sign up

Export Citation Format

Share Document